(Frontispiece?) 
CLAUDIAN AQUEDUCT (ROME), BUILT IN 50 A.D WATER-SUPPLY 
(CONSIDERED PRINCIPALLY FROM A 
SANITARY STANDPOINT) 
BY 
WILLIAM P. MASON 
Professor of Chemistry, RENSSELAE^^bLYTECHNic Institute 
Member of the A merican Philosophical Society, the A merican Chemical Society, the A merican 
Society of Civil Engineers, the American Institute of Chemical Engineers, the Ameri- 
can Public Health Association, the Royal Sanitary Institute {Great Britain), the 
American Water-Works Association, the New England Water-Works 
Association, the Franklin Institute, Honorary Member Association 
Generale des Hygienistes et Techniciens Municipaux (Paris), 
etc,, etc. 
FOURTH EDITION, REWRITTEN 
FIRST THOUSAND 
NEW YORK 
JOHN WILEY & SONS, Inc. 
London : CHAPMAN & HALL, Limited 
1916 Copyright, 1896, 1902, 1916 
BY 
WILLIAM P. MASON 
PRESS OF 
BRAUNWORTH Si CO. 
BOOK MANUFACTURERS 
BROOKLYN, N. Y. PREFACE 
So much materia] has been added to our stock of general 
knowledge upon the subject of 11 Water-supply " since the 
appearance of the last edition of this book that it has been 
found necessary to rewrite considerable portions of the original 
text, and to make many additions thereto. 
The writer hopes and believes that acknowledgment has 
been given whenever material has been extracted from either 
the published works or private notes of other authors, but he 
finds himself under obligations to so many persons, both here 
and in Europe, for courtesies extended and information fur- 
nished, that he is quite hopeless of properly recognizing them. 
William P. Mason. 
Rennselaer Polytechnic Institute, 
Troy, N. Y., November i, 1916. 
V CONTENTS 
CHAPTER I 
' PAGE 
Introductory i 
Magnitude of Ancient Water-supplies. Roman Aqueducts and the 
Supply of Rome. 
CHAPTER II 
Drinking-water and Disease io 
" Normal " and " Polluted " Waters. Peaty, Brown, and Swamp 
Waters. Odors and Tastes found in Waters. Wholesomeness of Hard 
Waters. Influence of Turbidity upon Health. Relation between Tur- 
bidity and Presence of Bacteria. Sewage-polluted Waters. Analyses 
of Sundry Epidemics of Cholera and Typhoid Fever. Relation of Ty- 
phoid-fever Death-rate to Improved Water-supply. Drinking-water of 
India and China. Power of Water to Carry Specific Disease. Viability 
of the Cholera and Typhoid Germs. Sterilizing Action of Sunlight. Ac- 
tion of Cold upon Bacteria. Statistics as to Sources of Typhoid Fever. 
Typhoid Fever and Rainfall. Estimated Yearly Tax Levied upon the 
Community by Typhoid Fever. Typhoid Fever Death Rates. Culti- 
vated and Uncultivated Typhoid Bacilli. Typhoid Bacilli in Car- 
bonated Waters, Sewage, Soil, Mud, Dust, Growing Plants, River 
and Lake Waters, Chicago Drainage Canal. Bacillus Coli Communis 
as an Indicator. Typhoid as a Country Disease. Typhoid Distributed 
by Milk, Flies, Shellfish, Secondary Infection. Typhoid and Filth. Re- 
sidual Typhoid. Tolerance. Attenuated Bacilli. Typhoid an Autumn 
Disease. Winter and Spring Typhoid. Relation of Typhoid to Height 
of Ground Water. Age of Greatest Fatality.- " Carriers." Diarrhoeal 
Outbreaks. Emergency Water Intakes. Danger Due to New Construc- 
tion. As to Using Distilled Water. 
CHAPTER HI 
Artificial Purification of Water 116 
General Necessity for Water Improvement. English Filter-bed System. 
Composition of Foreign and American Filter-beds. Analysis of Sand. 
Ice on Filters. Under Drains. Covers. Clear Water Basins. Cost 
VII VIII 
CONTENTS 
PAGE 
of Building and Maintenance of Filters. Efficiency of Filter-beds. 
Rates of Filtration. Methods of Cleaning Filters. Management of 
Filters. Mechanical Filtration. Anderson's Process. Filter-galleries 
and Cribs. Distillation. Aeration. Electrical Methods of Purifica- 
tion. Household Filtration. Charcoal Filters. Drifting Sand Filters. 
Puech-Chabal System. Non-submerged Filters. Chlorination. Use 
of "'Bleach " and Liquid Chlorine. Management of Swimming-pools. 
Purification by Ozone. Ultra-violet Light. Iron Removal Plants. Man- 
ganese Removal. Carbon Dioxide Removal. Emergency Purification 
Methods. Purification of Water for Army Use, 
CHAPTER IV 
Natural Purification of Water 221 
Nitrification. Sewage Purification at Asnieres. Intermittent Soil 
Filtration. Direct Oxidation. Sedimentation. Purification by Freez- 
ing. Purifying Action of Sunlight. Self-purification of Streams. Rate 
of Purification Varies with Amount of Contamination. Changes in 
Fresh Sewage upon Standing. Seasonal Variation in Purity of Streams. 
Laws Relative to Pollution of Streams. Chicago Drainage Canal Case. 
CHAPTER V 
Rain, Ice, and Snow 243 
Impurities in Air. Country and City Air. Country and City Rain. 
Monthly Variation in Composition of Rain-water. Impurities in Rain- 
water. Tanks and Cisterns. Ice as Food. Laws to Prevent Sale 
of Impure Ice. Viability of Bacteria in Ice. Ice and disease. Purifica- 
tion Due to Freezing. Improvement of Ice Due to Storage. Influence 
of Freezing upon Typhoid Bacilli. Ice-bome Epidemics. Strength of Ice. 
Anchor Ice. Ice Caves. Artificial Ice. Snow. Country and City 
Snow. Wholesomeness of Snow-water. 
CHAPTER VI 
River- and Stream-water 262 
Seasonal Variations in Composition of River-water. Discharge and 
Sediment of Rivers. Sewage Pollution of Rivers. Effect of Pollution 
upon Fish Life. Rainfall, Evaporation, and Flow of Streams. Normal 
Rainfall, by States, of the United States. Relation of Evaporation to 
Rainfall. Lines of Equal Evaporation for the United States. Rainfall 
and River-flow. Influence of Forests upon Water-supply. Erosion. 
Proper care of a Watershed. CONTENTS 
IX 
CHAPTER VII 
PAGE 
Stored Water 304 
Lake-water. Pollution of the Great Lakes. Chicago Drainage Canal. 
Advantages and Disadvantages of Reservoir Storage. Evidence of 
Sedimentation. Vertical Circulation in Lakes and Deep Reservoirs. 
The Stagnant Bottom Layer. Plankton. Cause of Coloring Matter 
and the Bleaching Action of Light. Changes in Ground-water during 
Open Storage. Growth of Algae in Stored Water. " Coppering " 
Reservoir Water. Amount, of Copper in Sundry Foods. Amount of 
Copper Sulphate to Kill Certain Algae. Effect of Copper Sulphate on 
Fish. Copper Sulphate as a Germicide. Crenothrix. Preparation of 
Reservoir Bottoms. Results of Stripping. Sedimentation in Reservoirs. 
Covered Reservoirs. Water Supply of Gibraltar. Dew Ponds. Care 
of Water Sheds. Disinfection of Reservoir. Effect of Street-main 
upon Bacteria in Water. 
CHAPTER VIII 
Ground-water 373 
Physical Properties of Soils. Movement of Water through Soils. 
Underground Streams versus Water-table. The " Underflow " of the 
Plains. General Character of Ground-water. Dug and Driven Wells. 
" Silting up " of Gang Wells. Infiltration-galleries. Pollution of 
Ground-water. Sub-surface Dams. Location of Wells. Contamina- 
tion by Privy Vaults. Reliance to be Placed upon Purification by 
Filtration through Soils. Relation of Typhoid Fever to Water and 
Drainage. Testing Wells for Possible Contamination. Pollution by 
Gasoline. 
CHAPTER IX 
Deep-seated Water 419 
Conditions Governing the Storing of Seep-seated Waters. Methods of 
Securing Deep-seated Waters. Drilled Wells. Diamond Drills. " Stove- 
pipe " Wells. Methods of its Reaching the Surface. Sea-springs. 
Artesian Wells. " Breathing Wells." Capacity of Rocks to Absorb 
Water. Exhaustion of Deep-seated Water. Copenhagen Supply. Char- 
acter of Deep-seated Water. Contamination of Deep-seated Water. 
Bacteria in Deep-seated Water. Sea-mills of Greece. 
CHAPTER X 
Quantity of Per Capita Daily Supply 449 
Statistics of Per Capita Supply in American and Foreign Cities. 
Statistics Showing Waste of Water. Influence of Meters in Preventing 
Waste. Influence of Meters upon Public Health. Estimated Future 
Population of Great Cities. Increase in Consumption of Water. Water 
for Fire Purposes. X 
CONTENTS 
CHAPTER XI 
PAGE 
Action of Water upon Metals 458 
Tanks, Pipes, Conduits, Boilers, etc. Action upon Lead, Iron, Zinc, 
Copper, and Galvanized Iron. Tuberculated Pipes. Cleaning of Same. 
Protection of Watermains. The Bower-Barff and Other Processes. 
Corrosion of Boilerplates. Boiler-scale. Boiler-scale " Preventives." 
Rusting of Iron. Boiler Waters. Water Softening. Permutit Process. 
" Red Water." 
APPENDIX A 
Analyses of Sea Waters 489 
APPENDIX B 
Rights and Duties Regarding the Pollution of Streams 490 
APPENDIX C 
Typhoid Fever Contracted from Drinking Polluted Water Decided 
to be "an Accident" 494 
Index... 497 WATER-SUPPLY 
CHAPTER I 
INTRODUCTORY 
From remote antiquity the highest value has been set upon 
an abundant and pure water-supply. Centres of population 
sprang up in ancient times around those points where it was 
readily available, and great expenditures of labor and treasuie 
were made to carry it to places where it was not naturally 
plentiful. Not only was a generous daily per capita allowance 
sought for, but we note in the centuries gone by unmistakable 
evidences of a keen appreciation of the dangers lurking in a 
polluted supply, and upon this point many of the ignorant 
consumers of our own day and generation would be benefited 
did they consult the wisdom of the past. 
Hippocrates, for instance, who wrote upon the value of 
pure water some four hundred years before the beginning of 
our era, advised boiling and filtering a polluted water before 
using it for drinking-advice which all must consider entirely 
" up to date." 
Pliny (a.d. 70) in his " Natural History " (book xxxi, 
chapters 1 to vi) devotes large space to the discussion of potable 
water, and thus speaks of one of the numerous supplies of Rome, 
which, by the way, is a water in use to-day: 
" Among the blessings conferred on the city by the bounty 
of the gods is the water of the Marcia, the cleanest of all 
the waters in the world, distinguished for coolness and 
salubrity." 
Libavius in 1595 refers to Pliny's work, -and adds the 2 
WATER-SUPPLY 
curious suggestion that the weight of a water is proportionate 
to its potability. 
During the Middle Ages it was observed that water some- 
times became poisonous through being distributed in lead pipes. 
Although Lascaris, who died in 1493, did not recognize 
the power of water to intensify and spread certain epidemics, 
it is interesting to observe that his teachings upon the origin 
of disease came very near the germ theory of the present day. 
The following from The Hospital is worth preserving as 
showing how our ancestors regarded the question of water- 
drinking : 
" It needed a very bold man to resist the medical testimony 
of three centuries ago against water-drinking. Few writers 
can be found to say a good word for it. One or two only 
are concerned to maintain that ' when begun in early life it 
may be pretty freely drunk with impunity,' and they quote 
the curious instance given by Sir Thomas Elyot in his 1 Castle 
of Health,' 1541, of the Cornish men, 1 many of the poorest 
sort, which never, or very seldom, drink any other drink, be 
notwithstanding strong of body, and like and live well until 
they be of great age.' Thomas Cogan, the medical school- 
master of Manchester fame, confessed in his ' Haven of Health,' 
1589, designed for the use of the students, that he knew some 
who drink cold water at night or fasting in the morning with- 
out hurt; and Dr. James Hart, writing about fifty years later, 
could even claim among his acquaintances ' some honor- 
able and worshipful ladies who drink little other drink, and yet 
enjoy more perfect health than most of them that drink of 
the strongest.' Sir Thomas Elyot himself is very certain, 
in spite of the Cornish men, that there be in water causes of 
divers diseases, as of swelling of the spleen and liver. He 
complains oddly, also, that ' it fitteth and swimmeth,' and con- 
cludes that 1 to young men, and them that be of hot complexion 
it does less harm, and sometimes it profiteth, but to them that 
are feeble, old, and melancholy it is not convenient.' But the 
most formal indictment against water is that of Venner, who, 
writing in 1622, ponderously pronounces to dwellers in cold INTRODUCTORY 
3 
countries ' it doth very greatly deject their appetites, destroy 
the natural heat and overthrow the strength of the stomach, 
and, consequently confounding the concoction, is the cause of 
crudities, fluctuations, and windiness in the body.' " 
Much the larger undertakings connected with ancient 
water-supply were those built entirely for, or at least in con- 
nection with, general systems of irrigation. " The extent of 
some of these great hydraulic works can be conjectured from 
the ruins remaining. Lake Moeris, in Egypt, was constructed 
at least 2000 years before Christ. Its dimensions were suffi- 
cient to regulate the annual inundation of the Nile, receiving 
the surplus waters when there was danger of a flood, and sup- 
plying the needed deficiency when the river reached a stage 
which could not irrigate the crops. This, with other large 
reservoirs of flood-waters, enabled a population of 20,000,000 
to exist in the valley of the Nile, while it now supports barely 
one-fourth of the number. 
" In ancient times the valleys of the Euphrates and Tigris, 
now almost a desert, were densely populated. Four thousand 
years ago the rulers of Assyria had converted those sterile 
plains and valleys into gardens of extreme productiveness by 
the construction of immense artificial lakes for the conserva- 
tion of the flood-waters of the river, and great distiibuting- 
canals for irrigation. One of these canals, supplied by the 
Tigris, was over 400 miles long and from 200 to 400 feet broad, 
with sufficient depth for the navigation of the vessels of that 
time.* 
11 In India tanks, reservoirs, and irrigating-canals were 
constructed many centuries before the Christian era, and a 
great part of that country was kept in the highest state of 
cultivation. Some of the tanks or artificial lakes covered many 
square miles, and were often fifty feet in depth. 
" Evidences exist in New Mexico and Arizona that in pre- 
historic times a race now extinct had extensive irrigation- 
works and cultivated large areas." (Wyckoff.) 
* The Nahrawan Canal. It is of great antiquity, and its former importance 
is shown by the ruins situated along its course. 4 
WATER-SUPPLY 
Professor F. H. Cushing advises the author of his dis- 
covery of ancient reservoirs of large size in southern Arizona. 
On the site of ancient Carthage there are still to be seen 
THE CISTERNS AT CARTHAGE. ALL THAT IS LEFT OF THE ANCIENT CITY. 
the great cisterns for storage of water, eighteen in number, 
and each about one hundred feet long, twenty feet wide, and 
nearly twenty feet deep. They were originally covered with INTRODUCTORY 
5 
earth, and to-day are in marvellously good repair. They lie 
in two long rows and empty into a common gallery situated 
between the rows. They belonged to the ancient or Punic 
city. * 
Drinking-water was supplied to ancient Carthage from a 
spring in the Zaghoun Mountains some sixty kilometres to 
the south. The channel way, which the writer found to be 
only ten inches square in section, contoured the hillsides for 
considerable distances, at times went under ground, and on 
approaching the coast was carried on arches of a magnitude 
seemingly out of proportion to so small a channel. From 
ten to fifteen kilometres of the old aqueduct yet remain. See 
illustration, page 6. 
" Amongst the wonderful monuments of the former great- 
ness of the Singhalese people must be mentioned the ruined 
tanks, with which scarcely anything of a similar kind, whether 
ancient or modern, can be compared. Thirty colossal reser- 
voirs and about seven hundred smaller tanks still exist, 
though for the most part in ruins. In February, 1888, the 
largest and most important tank in Ceylon, that of Kalawewa, 
was, after four years of labor, completely restored. It was 
built 460 a.d. to supply Anuradhapura with water, but has 
been in ruins for centuries. Now again it contains an area of 
seven square miles of water, twenty feet deep, and supplies 
smaller tanks more than fifty miles distant." f 
When we reflect that these great works of antiquity were 
accomplished without the aid of steam, electricity, or explo- 
sives, we are impressed with the belief that, in intelligent 
perseverance at least, " there were giants in those days." 
Our respect does not lessen when we contemplate the ex- 
tent of the supply of water poured into the " Eternal 
City." 
The following is freely taken from Forbes's lecture on the 
Roman aqueducts: 
* See " Carthage and Carthaginians," by Bosworth Smith, 
f Chambers's Encyclopaedia, iii. 80. 6 
WATER-SUPPLY 
AQUEDUCT NEAR TUNIS, LEADING TO ANCIENT CARTHAGE. INTRODUCTORY 
7 
Date of Construction. 
Length. 
Appia 
. . . 312 B.C. 
11 miles 
Anio Vetus 
272 ' ' 
43 " 
Marcia 
... 145 " 
61 " 
Herculea branch.. . . 
3 11 
Tepula 
126 " 
13 " 
Julia 
■■■ 34 " 
15 " 
Virgo 
21 " 
14 11 
Augusta 
. . . IO A.D. 
6 11 
Alsietina 
IO " 
22 ££ 
Claudia 
... 50 " 
46 ££ 
Anio Novus 
... 52 " 
58 " 
Neronian branch. . . 
97 " 
2 ££ 
Traiana 
109 to 200 A.D. 42 miles 
Hadriana 
••• 117 t0 x585 
" 15 " 
Aurelia 
162 
C< l6 ££ 
Severiana 
200 
££ 10 ££ 
Antoniniana branch. 
215 
" 3 " 
Sabina-Augusta. . . 
130 to 300 
" I5 " 
Alexandrina 
226 
" 15 11 
ROMAN AQUEDUCTS, ARRANGED IN CHRONOLOGICAL ORDER 
(The miles above given are Roman, of 4854 feet. The entire length of the 
aqueducts in English miles would be 381.) 
It has been calculated that, altogether, the supply was 
332,306,600 gallons daily, which would have been over 332 
gallons per capita upon a basis of a population of one million. 
This calculation, however, has been based upon very indif- 
ferent data furnished by Prony in 1817. As Mr. Clemens 
Herschel has lately shown, a much more probable figure for 
the daily water consumption of the world's capital should read 
thirty-two million U. S. gallons. 
Notwithstanding that some of the Roman aqueducts were 
damaged during the wars of the sixth and seventh centuries, 
the supply did not entirely cease until the fourteenth century, 
when Rome was abandoned by the papal court. In the present 
day four of the ancient sources still supply the city with water. 8 
WATER-SUPPLY 
Of the imposing lines of arches which stalk across the 
Campagna none is more interesting than the stately Claudian 
aqueduct. (See frontispiece.) 
Speaking of it, Pliny says: 11 The preceding aqueducts 
have all been surpassed by the costly work more recently 
commenced by the Emperor Caius and completed by Claudius. 
The sum expended on these works was 350,000,000 of ses- 
terces." * 
Near the city the Claudian arches were filled up by Aure- 
lian, and made to do duty as part of the city wall, the great 
gateway permitting passage at this point being known as 
the Porta Maggiore. 
It is curious to observe that waters from different sources 
were carried in separate channelways upon the same arches. 
Speaking of the Marcia, Tepula, and Julia waters, Fron- 
tinus says: 11 The three are carried on the same arcade, the 
highest being the Julia, then the Tepula, and lowest the 
Marcia." Near the Porta Maggiore the three channels are 
still distinctly to be seen. 
It remains to say a word about the " castella " so often 
found along the courses of the Roman aqueducts. Forbes 
calls them " filtering-places," but such they could not have 
been-at least in the modern acceptation of the expression. 
They varied in size and in the number of their chambers, 
some having but four, while others numbered as many as 
twelve compartments. One of the smallest size is illustrated 
on p. 9. 
The water entered chamber A and passed by means of 
holes in the floor into C, thence through openings in the wall 
into D, from which it rose through holes in the floor into B, 
and then passed on to Rome. A breaking of undue 11 head " 
would take place in the 11 castellum," as is accomplished in a 
more modern fashion at Vienna to-day; f but the real benefit 
* About $12,700,000. 
t At Vienna the aqueduct has a fall of 11,000 feet in the first 10 miles, and 10 
feet per mile the remainder of the way. It is about 60 miles long. The " head " 
is throttled by gates, every 100 feet of fall on the upper section. See Engineering 
News, Nov. 29, 1894. INTRODUCTORY 
9 
derived from the construction of these chambers was probably 
the opportunity given for sedimentation. 
Thus Frontinus says: " At the seventh mile, on the Via 
Latina, the Marcia, Tepula, and Julia are taken into a covered 
ANCIENT ROMAN 
CASTELLUM 
' filtering-place,' where, as though breathing again after their 
course, they deposit mud." 
So far as quality of water is concerned, it is probably true 
that the ancient public supplies were better than the average 
of what is furnished to our cities to-day, for the reason that 
pumping was unknown and the water-purveyors were of neces- 
sity driven to distant and purer sources to the exclusion of 
local rivers which lay too low to be utilized. CHAPTER II 
DRINKING-WATER AND DISEASE 
In that excellent treatise upon " Water-supplies and Inland 
Waters " issued by the Massachusetts State Board of Health, 
waters are classified as " normal " and " polluted," the former 
being such as are free from addition, directly or indirectly, 
of either sewage or industrial waste. 
The relation of " normal " waters, as a class, to sanitary 
science constitutes a subject by itself, and one shrouded in much 
confessed ignorance and conflicting testimony, as is instanced 
by the doubt we entertain of the effect of " peaty water " 
upon the human organism. 
Bog-waters in Ireland, especially those of Lough Sheighs 
in the county of Cavan, and of Lough Neagh, have long been 
used for the treatment of skin-diseases. The mineral waters 
of Askern, in Yorkshire, England, are about saturated with 
peaty material from the neighboring bogs, and have for many 
years been successfully used in the treatment of chronic 
rheumatism and skin-diseases. 
Bothamley thinks it not improbable that the dissolved 
peaty matter is the curative agent in these waters.* 
" Certain affluents of the Orinoco and Amazon have what 
are called black waters {aquas negras}. When seen in mass, 
they are of a coffee-brown or of a greenish black. In the 
shade they are almost black, but in a glass they are brownish 
yellow, though very transparent. They have no disagreeable 
taste, and are preferred for drinking. 
" The waters do not undergo any chemical change on 
keeping." f 
* J. Chern. Soc., Ixiii. 696. 
f Chem. News, Iviii. 305. 
10 DRINKING-WATER AND DISEASE 
11 
This ability to " keep well " is very frequently observed in 
brown waters, but it is by no means universal, nor does it 
seem to be a necessary characteristic indicating potability. 
In writing to the author regarding the water of the Dismal 
Swamp of Virginia, Surgeon-General J. R. Tryon, of the 
U. S. Navy, said: 
"I beg to enclose herewith copy of analysis made of Lake 
Drummond (Dismal Swamp) water in October, 1891, at the 
Naval Museum of Hygiene. 
" This has always been considered a very pure water, and 
before the general adoption of condensers was much used for 
naval vessels. The bureau has no special data bearing upon 
the relation to disease with the use of the class of waters 
referred to," i.e., " peaty " waters. 
ANALYSIS OF LAKE DRUMMOND WATER 
(Parts per Million.) 
Color Yellow 
Odor Sweetish 
Turbidity None 
Sediment ; Very slight 
Residue on evaporation 294.0 
Loss on ignition ' 258.0 
Fixed solids 36.0 
Free ammonia .016 
Albuminoid ammonia 1.75 
Nitrogen as nitrites None 
Nitrpgen as nitrates 5.00 
Chlorine 1.50 
Hardness , 49-3 
(For data in general regarding the analysis of water, inter- 
pretation of the same, method of taking samples, methods of 
analysis, etc., reference may be made to the author's book 
on Examination of Water.') 
Many of the surface-waters of Massachusetts are colored 
brown, especially that of the Acushnet River, supplying the 12 
WATER-SUPPLY 
town of New Bedford. Experiments made with this water 
showed that its dissolved nitrogenous matter remained per- 
manent for months without the development of " free am- 
monia," or other indications of decay.* 
As illustrating the undesirable character of some swamp- 
waters, the celebrated case of the ship Argo may be called to 
mind; an incident too often recited to bear repetition here.f 
It has been the author's fortune to meet with but few 
cases of illness traceable to peaty waters, and in all such in- 
stances the patients suffered from a mild and transient form of 
diarrhoea, caused by water from a low-lying, shallow lake or 
pond, surrounded by low wooded banks. Whatever be the 
morbific principle of such waters, it appears to be removed by 
suitable filtration. 
Mrs. E. H. Richards very properly points out that such 
peaty waters as are found unwholesome may owe their toxic 
qualities to the presence of materials other than the brown 
coloring matter. • 
When we dwell upon the fact that the milder enteric dis- 
orders rarely get into the " death-rate," and that visiting 
strangers may suffer from a cause to which the acclimated 
natives are not susceptible, we appreciate that in dealing with 
peaty water we must consider it as largely an unknown 
quantity, probably entirely harmless, but yet the possible centre 
of trouble, especially if the amount of organic matter present 
be large. Even the same water may not at all times be equally 
potable, for, as J. W. Mallet has well said: " It seems quite 
conceivable that a water containing organic matter of any 
kind may be harmless at one time and harmful at another, 
when perhaps a different stage of fermentation or putrefactive 
change may have been entered upon, and special organisms 
may have made their appearance or entered upon a new phase 
of existence. Thus there might possibly be safety in drinking 
a peaty water, or water filtered through beds of dead forest 
leaves, when fresh; danger when, after a certain amount of 
* Mass. Bel. Health, 1890, 547. 
f Parke's " Hygiene." DRINKING-WATER AND DISEASE 
13 
atmospheric exposure, bacterial organisms had become de- 
veloped; and safety, again, perhaps, after the growth of such 
organisms had fallen off and more or less of the available 
organic matter had been consumed." 
When, therefore, the question arises as to the advisability 
of introducing a peaty water as a town supply, the possibly 
unsatisfactory character of the same must always be borne in 
mind, and the municipal authorities should be prepared for the 
probable necessity of constructing a filtering-plant. More- 
over, the likelihood of a filter-plant being required will increase ' 
as the years go by, for the people are fast inclining towards a 
demand for a clear and bright water as well as one that is 
potable. Furthermore, a distinction should be always drawn 
between a flowing water carrying fresh peaty material in solu- 
tion and a water derived from a stagnant swamp. " In 
certain cases it may be a fact that the algae decay rapidly 
but that new growth absorbs the products of decomposition, 
so that they do not accumulate in the water. Shallow, stag- 
nant bodies of water, which in the heat of summer are full 
of vegetable and animal life, become in time foul, because decay 
gets, ahead of growth and the products of decomposition 
accumulate." 
11 Malaria and Drinking-water " was the title of papers 
formerly found in our sanitary press. 
Surface waters and those from shallow wells in " malari- 
ous " districts were accused of producing the fever by an accum- 
ulation of testimony that was at the time difficult to resist; 
but the matter has been thoroughly investigated and we now 
know that " malaria " is due to the bites of infected mosqui- 
toes and that water is not a cause. 
Odors which at times occur in water, and which are vari- 
ously described as " musty," " fishy," " cucumber," " green- 
corn," " horse-pond," and the like, are produced by the 
growth, death, and decay of minute organisms, especially 
those known as algae. However objectionable these odors 14 
WATER-SUPPLY 
and tastes may be from an aesthetic standpoint, it has not been 
proven that they are directly productive of disease. 
D. D. Jackson has shown * that the odors caused by 
living microscopical organisms are due to compounds of the 
nature of essential oils; and Whipple points out that the 
amount of such oil produced by an abundant growth of the 
organisms is quite sufficient to account for the effect observed.! 
He notes, for comparison, that oil of peppermint can be 
recognized when diluted with water in the proportion of one 
part of oil to fifty million parts of water; and that when Asterion- 
ella is present to the extent of fifty thousand organisms per 
cubic centimetre the dilution of its oil is in the proportion 
of about one part to two million parts of water. 
Whipple further suggests that the flow of water through 
pipes may cause disintegration of organisms with liberation of 
the odor-producing oil; hence the odor at the tap may be 
greater than at the intake. May it not be, however, that the 
" tap " odor is of mixed quality, partly due to products of 
decomposition? 
This matter of odor and taste is again considered under 
" stored water," page 318. 
The droplets of oil above referred to are shown in the fol- 
lowing " camera lucida " drawing of a diatom. The drawing 
was made by Mr. W. H. Long.! 
Although natural waters are to be found bearing in solution 
large quantities and varied assortments of mineral ingredients, 
yet such cannot be considered potable, except in the sense that 
that are medicinal, and they consequently do not find place in 
* Tech. Quarterly, x. 419. 
t Microscopy of Drinking-water, p. 195. 
$ Bui. Univ. Texas, v. 22. DRINKING-WATER AND DISEASE 
15 
the present writing. When those minerals are present, however, 
which give to the water the characteristic known as hardness, 
the case is quite different, for such a property obtains, to a 
greater or less degree, in every public supply. As to the 
wholesomeness of such a 'water, there is a widespread opinion 
that a high degree of hardness is not compatible with safety; 
but although hard waters do often produce certain intestinal 
deiangements, notably constipation, in persons not accustomed 
to their use, there are no sufficient statistics on record tending 
to confirm the popular belief that their ingestion leads to the 
formation of urinary calculi. It is a matter of common obser- 
vation that sudden change from the use of a pure, soft water 
to one equally pure, but harder, or vice versa, results in tempo- 
rary intestinal derangement, showing the difficulty to be due 
rather to the change than to the inherent character of the water 
employed. While considering the wholesomeness of hard water 
the English Rivers Pollution Commission (Sixth Report) col- 
lected the following statistics, the comparison having been made 
between towns of the same class, in which the general conditions 
of life are similar. The conclusion was: " Where the chief 
sanitary conditions prevail with tolerable uniformity, the rate 
of mortality is practically uninfluenced by the softness or 
hardness of the water." 
Kind of Town. 
X 
Number 
of Towns. 
Average 
Population. 
Average 
Annual Mortality 
per 
1000 Inhabitants. 
Seaport 
5 
162,801 
29.4 
Inland manufacturing 
172,860 
20 7 
Inland non-manufacturing 
8 
10,751 
25-4 
Watering-places 
5 
48,430 
iQ-5 
TOWNS SUPPLIED WITH SOFT WATER 
TOWNS SUPPLIED WITH MODERATELY HARD WATER 
Seaport 
3 
226,172 
321 
Inland manufacturing 
8 
108,715 
62,372 
33480 
26.9 
Inland non-manufacturing 
4 
3 
Watering-places 
19.2 
.. 16 
WATER-SUPPLY 
TOWNS SUPPLIED WITH HARD WATER 
Seaport 
6 
116,406 
25-1 
Inland manufacturing 
5 
144,981 
255 
Inland non-manufacturing 
20 
29,169 
23.2 
Watering-places 
12 
53,170 
20.4 
More recently Dr. J. C. Thresh, Medical Officer of Health 
for the county of Essex, England,* gave statistics showing 
that, in the areas supplied with hard water, the death-rate is 
as low, or lower, than in those furnished with soft-water. 
He concludes that " the character of the water supplies in the 
county had no effect upon the death-rates." His figures, 
which follow, indicate the average annual deaths per 1000 
population for the years 1907-1911 inclusive; and for 1912 
separately. 
Death-rate 1907-n. 
Death-rate 1912. 
Aver- 
age. 
Extremes. 
Aver- 
age. 
Extremes. 
Max. 
Min. 
Max. 
Min. 
Soft water area 
II. 0 
II . 2 
8.9 
10.0 
U.S 
8-7 
Moderately hard water areas: 
Country towns 
115 
12.8 
7-9 
10.9 
14.0 
7-6 
South Essex 
II . 2 
13-0 
8.8 
9.9 
10.6 
6.2 
Hard water area: 
Country towns 
II . 2 
14.I 
8.0 
9.9 
12.9 
7-i 
Metropolitan 
II. I 
12 . I 
8-5 
10.6 
II. I 
7.2 
The " hardness " referred to he describes as: 
" Soft," under 10 degrees, i.e., 143 parts per million. 
" Moderately hard," from 10 to 20 degrees, i.e., 143 to 286 
parts per million. 
"Hard," from 20 to 30 degrees, i.e., 286 to 429 parts 
per million. 
These are English notions of " hardness." In the United 
States we classify waters as: 
" Soft," 50 or less parts per million of " hardness." 
" Hard," 100 or more parts per million of " hardness." 
" Doubtful," between the above figures. 
* Water and Water Engineering, April 15, 1914. DRINKING-WATER AND DISEASE 
17 
Dr. Allfrey practiced medicine a great many years in Kent 
(England) and was familiar with the very hard water of the dis- 
trict. He states: " The hardness was an intolerable nuis- 
ance," but he " had not observed that it was in any way in- 
jurious to health, or predisposed in any way to gouty ailments 
as was popularly supposed." * 
The following Massachusetts data collected by A. L. Gam- 
mage f entirely support those already given and indicate no 
relation between hardness and health. 
RELATION BETWEEN WATER HARDNESS, TYPHOID AND TOTAL 
DEATH RATES IN MASSACHUSETTS 
Ground Waters-34 Cities and Towns. 
Average Hard- 
ness, Parts per 
1,000,000. 
Death Rates per 1000. 
Typhoid. 
Total. 
Softest 
17 
0.08 
14.18 
Medium 
29 
0.14 
I5-7I 
Hardest 
71 
0.11 
14.12 
Surface Waters-39 Cities and Towns. 
Softest 
5-7 
0.115 
14.99 
Medium 
17 
O. IO 
1385 
Hardest 
41 
0.14 
14 98 
Production of goitre through the drinking of certain waters, 
particularly those derived from the melting of ice or snow, 
has been widely commented upon. For instance: 
" In Bozel there was a population of 1472 people. Of this 
population no fewer than 900 had goitre and 109 were cretins. 
The water supply was hard and contained sulphates of cal- 
cium and magnesium. The authorities laid on a soft water 
and sixteen years later the number of persons suffering from 
goitre was reduced to 39 and the number of cretins to 58." J 
According to Dr. Rupert Farrant, surgical registrar to the 
Westminster Hospital, " Goitre is common in the valleys of 
mountainous districts. The water supply in these districts 
* J. Royal San. Inst. 26: 682. 
t Engineering News, December 2, 1915. 
I J. Royal San. Inst., 34: 299. 18 
WATER-SUPPLY 
is of surface origin, and so liable to surface pollution, as it neither 
passes through filtering beds nor stands in volume. Water 
derived from snow is open to surface contamination, and, 
as shown by Houston, micro-organisms survive longer in 
water at lower than at higher temperatures. The coll, being 
placed under abnormal conditions, either in the water or in 
the upper part of the intestinal tract, undergo mutation. The 
mutants are present in the faeces of individuals affected with 
endemic goitre, and when once lodged there may remain for 
many years. Dr. Farrant concludes that endemic goitre is 
preventable by the avoidance of water contamination and by 
the sterilisation of contaminated water."* 
Apparently the filtration of water has no effect on the goitre- 
producing substance in it. This has been demonstrated by 
Bircher, who 11 gave to monkeys, imported from Africa to 
Hamburg, the water of a certain locality where goitre is 
endemic. They all showed enlarged thyroids. He tried the 
same water, after filtration, with like results. Monkeys which 
were given this water boiled showed no thyroidal enlarge- 
ments." f 
In an article on 11 The Importance of Magnesia in Drink- 
ing-water " J Percy Frankland shows that a very great per- 
centage of the population of England are to-day using waters 
containing from 40 to 60 parts of MgO per million; and that 
consequently the prejudice existing against magnesian waters, 
on account of their supposed production of calculi, goitre, and 
cretinism, should not be given undue importance. 
In 1889 Professor Kocher investigated the occurrence and 
distribution of goitre in the canton of Bern, Switzerland, and 
published an interesting map showing the varying frequency 
of the disease with change of geological conditions. § He was, 
moreover, able in a few instances to definitely classify certain 
local waters as producers or non-producers of goitre. The 
* Water and Water Engineering, May 15th, 1915. 
f Albany Med. Anal., Dec., 1914. 
J International Congress of Hygiene, London, 1891. 
§ Vorkommen und Vertheilung des Kropfes im Kanton Bern. DRINKING-WATER AND DISEASE 
19 
analyses of two such waters are here given, in parts per 
million: 
Producing 
Goitre. 
Not producing 
Goitre. 
Specific gravity at 170 C 
1.000316 
I.0004I 
Lime (CaO) 
79 • 
80 7 
Magnesia (MgO) 
5 • 4 
10 9 
Calcium sulphate (CaSO4) 
9.01 
36 I 
Chlorine (Cl) 
4 • 97 
7. 
Silica (SiO2) 
3-3 
. 4-5 
Iron and alumina 
traces 
traces 
Organic matter 
( c 
I c 
Nitrates 
t c 
c I 
Nitrites 
none 
none 
Ammonia ; 
( c 
c c 
Iodine and bromine 
c c 
c I 
It is interesting to note in the above analyses that the 
magnesia, about which fears have been expressed, is much 
larger in the non-goitre-producing water than in that which is 
credited with causing the disease. 
Professor Kocher has made cultures of the bacteria occur-' 
ring in the two classes of water, and has injected rabbits there- 
from, with the result that glandular swellings have been pro- 
duced in several of the animals treated with preparations 
from the 11 goitre-water." Nothing definite has, however, 
been as yet secured, and our present knowledge upon the 
subject is in a very unsatisfactory state. 
Philippe,* after commenting upon the various theories 
advanced to account for the occurrence of goitre, concludes 
that 11 practically nothing is known as to the true cause of the 
disease." 
An article, with map, prepared by Dr. Thierry on " Goitre, 
and the question of supplying Paris with water from Lake 
Leman," will be found in 11 L'Ean " for September 15, 1913. 
Lying between ordinary hard waters and those of a true 
mineral character is a group, principally of artesian origin, 
* Mitt. Lebensm. Hy., 5, 1. 20 
WATER-SUPPLY 
containing sundry objectionable mineral ingredients, such 
as magnesium chloride, hydrogen sulphide, and the like, and 
which would hardly be considered potable did they occur 
in well-watered regions, nevertheless they are thankfully 
received by people very willing to take almost any water they 
can obtain. 
Turbidity is exceedingly common in the river-waters of 
this country, particularly in those of the great central basin. 
The quantity of suspended claylike material present naturally 
varies in each stream with the conditions of the season, being 
at times entirely absent in some of them, while with others 
the existence of more or less muddiness appears to be a con- 
stant characteristic. 
Ockerson reports that four different surface-samples of 
Mississippi river-water contained 576, 788, 1030, and 348 
parts per million of sediment. On July 2, 1894, the same 
river-water, at New Orleans, contained 2360 parts per million 
of solids, mostly silt. Below the junction of the clear Missis- 
sippi with the muddy Missouri, the two waters flow on for 
miles, side by side, with a. distinct line of separation. 
With reference to the influence of the suspended mineral 
matter upon health, we find some conflict of opinion. It is 
an unquestioned fact that very roily water is ingested by 
many of our communities with no traceable evil results, but the 
preponderance of testimony goes to show that immunity is 
attained by continued use, and that the visiting stranger is 
not upon the same platform of safety with the acclimated native. 
In reply to the claim so often advanced that turbidity is 
a positive advantage, as tending to remove objectionable 
material from a sewage-polluted river-water, it should be 
stated that suitable arrangements for sedimentation must be 
furnished, otherwise no advantage can be expected from the 
mere presence of the suspended mineral ingredients. It is a 
well-known fact that precipitating solids will drag down with 
them other finely divided substances which, if left to them- 
selves, would require long periods of time for complete sedi- DRINKING-WATER AND DISEASE 
21 
mentation, and that even soluble salts may be in part carried 
down by the same cause. 
It may readily be conceived that, acting in obedience to 
this principle, the depositing silt would gather to itself, and 
carry with it, many germs of disease which, if left to them- 
selves in clear water, would require much longer time to fall; 
but that there is any advantage to be looked for in using a 
turbid water without sedimentation, and thereby swallowing 
turbidity, germs, and all, is scarcely rational. 
The following is in illustration of the influence of turbidity 
in causing a more rapid deposition of bacteria diffused through- 
out a body of water. The results represent the relative num- 
bers, per cubic centimetre, at the surface of a body of water, 
and at varying depths in the same, under the condition of 
clearness and also under that of material turbidity. 
The experiment was conducted with a tall tin vessel, one 
foot in diameter, and tubulated at intervals of one foot, to 
allow drawing of samples. The time of settlement was made 
two hours; and the numbers of bacteria found per cubic centi- 
metre at the successive depths are stated in terms of the num- 
ber in the surface sample, that being called one hundred. 
Muddy Water 
Clearer Water. 
Surface 
IOO 
Depth of one foot 
134 
125 
Depth of two feet 
166 
142 
Depth of three feet 
186 
169 
Depth of four feet 
266 
254 . 
Two hours of settlement were not enough to bring out 
marked contrast between the waters, although the principle 
was sustained. 
The really serious item of contamination, the one to which 
the sanitarian's attention is most quickly drawn, is that of 
sewage pollution, and a consideration of the questions arising 
upon this topic dwarfs all others into comparative insignifi- 
cance. 22 
WATER-SUPPLY 
" To those of us who never had our attention called to the 
fact, it may be something of a revelation to know the amount 
of waste necessarily disposed of in our great cities. For in- 
stance in London this waste has been estimated closely for 
1000 people at 260 tons per annum, or over 2| tons per person 
per annum. This estimate, of course, includes all characters 
of waste, like ashes, papers, cans, etc., as well as sludge, but 
omits the vast amount of water separated from the sewage 
in forming the sludge. Pettenkoffer estimates 90 grams of 
feces and 1170 grams of urine as the average excrement per 
diem for men, women and children. This, coupled with a 
minimum allowance of 159 liters of water for household 
use, would make a total of 160 liters, or something over 35 
gallons of sewage, approximating 280 pounds each day per 
person." * 
The city of New York pours something like six hundred 
million gallons of sewage into the Hudson River and New York 
Bay every twenty-four hours, and, upon the completion of the 
Passaic Valley trunk sewer, about fifty per cent more will be 
added from cities and towns in New Jersey. This does not 
include garbage, which is disposed of by other means. 
Shall a water once polluted with sewage material be again 
used for human consumption? If there be danger in such use, 
what is its nature, what is its extent, and are there available 
means of averting it? These are popular questions of the 
day with which the sanitarian has to grapple. 
It would hardly be wise to take the reader's time with a 
resume of matters, possessing only historic interest, pertaining 
to water pollution; suffice it to say that, within the very 
recent past, strong views were entertained concerning the self- 
purification of streams, and also upon certain features of natural 
and artificial filtration, which we now believe to have been 
erroneous. That polluted water-supplies have caused wide- 
spread illness and death is established beyond a peradventure, 
and, from among the many illustrations that might be cited, 
* N. Y. State J. of Medicine, July, 1910. DRINKING-WATER AND DISEASE 
23 
the author offers the following in evidence, some of the data 
having been collected by himself or within his personal ex- 
perience : 
In the autumn of 1887 the city of Messina, Sicily, was 
visited by an epidemic of cholera. The plague lasted from 
September 10th to October 25th, during which time there 
were some 5000 cases and 2200 deaths. Although for a time 
the number of daily cases was excessive, running as high as 400, 
the ordinary number was about 70. The population was 
stampeded, falling from 71,000 to about 25,000. The govern- 
ment felt that a very possible cause for the rapid spread of the 
scourge lay in a contaminated drinking-water, and an inquiry, 
resulting in a development of the following facts, fully con- 
firmed the suspicion: The water as it left the gathering- 
grounds in the mountains was of excellent quality, but it was 
conveyed to the city in a conduit entirely open. Those who 
are familiar with European customs will remember that the 
washing of soiled clothing is there largely an out-of-door occu- 
pation, conducted in the nearest available watercourse. For 
the benefit of the Messina washerwomen a portion of the public 
water was deflected, before reaching the town, and turned into 
neighboring washing-pools of stone. A fair proportion of this 
deflected water, after having been used for laundry purposes, 
found its way back into the channel, and continued its course 
to the city. Further contamination occurred within the town 
itself, for the reason that the mains of the distributing system 
were of unglazed tile, badly joined, and were laid in the 
immediate vicinity of unglazed tile sewers, also very leaky. 
The sewers were at times found on top of, and parallel with, 
the water-mains. 
Acting upon its conviction as to the cause of the great 
mortality, the government sent tank-ships to the mainland, 
filled them with pure " Serino " water, supplied the people 
therewith, and the daily number of cholera cases immediately 
fell from seventy to five; or, to quote an expression of the 
time, " the plague ceased as if by magic." A more efficient 
distributing' system has since been introduced, the open con- 24 
WATER-SUPPLY 
POLLUTION OF A STREAM BY THE WASHING OF SOILED LINEN THEREIN-GENEVA, SWITZERLAND. DRINKING-WATER AND DISEASE 
25 
duit has been replaced by modern pipe, and the city has escaped 
further visitation by cholera. 
From what the author saw, however, it would seem that 
there is much yet to be done towards bettering the Messina 
water-supply. Great tanks still exist in the suburban gardens 
where laundry waste-water is stored for irrigation purposes. 
Water from wells in these gardens is taken by the city 
authorities during periods of scant supply from the regular 
source, thus permitting certain pollution of the municipal 
supply. 
The influence of the washerwomen in spreading cholera in 
Messina reminds us that the great epidemic at Cuneo, Italy, 
in 1884, resulting in 3344 cases, was traced to identically the 
same source. Infected linen had been washed in a brook 
communicating with the public water-supply. 
In 1890 two violent outbreaks of typhoid fever occurred in 
the valley of the Tees River.* 
The Tees in a small stream of northern England, about 
seventy miles long, and navigable for about four miles from 
its mouth. 
Most of the towns of the valley take their water-supplies 
from the river, but a large scattered population receives water 
from other sources. The estimated population using Tees 
water at the time of the outbreak was 219,435, and the num- 
ber not using such water was 284,181. 
In many places, especially in the towns, the river receives 
all sorts of polluting additions, which are carried on by the 
current to the intakes below. During dry weather the stream 
recedes considerably, leaving uncovered its rocky foreshores, 
which accumulate filth of every variety, and retain the same 
until, by reason of heavy rain, the river suddenly rises and 
sweeps the refuse onward toward the towns nearer the sea. 
The result produced upon the thoughtless public of such 
an extra and concentrated dose of sewage material added to 
their water-supply is best shown graphically by the accom- 
* See 21 st Report of the London Local Government Board. 26 
WATER-SUPPLY 
panying chart, where it will be observed that increase of rain- 
fall is followed by increase in cases of typhoid fever among the 
RAINEALL(BARNARD castle) 
SNOW REPRESENTED BY SHADING 
v/A va/vza tv* iii' r-n rill wn-vzn-wa - - 
RAI N FALL (BARNARD CASTLE) 
SNOW REPRESENTED BY SHADING 
typhoid fever rates 
PEP IQ.000 POPULATION 
TYPHOID FEVER RATES 
PER 10,000 POPULATION 
I1890■ 1891 
A PERSONS USING TEES WATER 
POPULAT ION(l89l) 219.435 
1890 1891 
B PERSONS NOT USING TEES WATER. 
POPULAT I 0N(l89l)284,l8J 
RAINFALL AND TYPHOID FEVER IN THE TEES VALLEY, ENGLAND. 
219,435 persons using the Tees water, after an interval corre- 
sponding to the incubation period of the disease, while no DRINKING-WATER AND DISEASE 
27 
appreciable result is noticed among 284,181 people of the 
same district, using other sources of supply. 
The 11 typhoid rates " given in the chart are " cases," not 
" deaths." 
It fell to the author's duty to investigate certain points 
relating to the typhoid epidemic occurring in the valley of the 
upper Hudson during the autumn and winter of 189'0-91, and 
the facts secured seem especially instructive. 
By a glance at the accompanying chart, the locations will 
be observed of the several cities and towns situated at and near 
the junction of the Mohawk and Hudson rivers. Every one 
of these centres of population drains into the river on whose 
bank it is situated, and each of them, except Lansingburgh, 
took at that time all or the greater portion of its water-supply 
directly from such river by means of pumps. Mark this dif- 
ference, however, Waterford and Troy were supplied with 
Hudson River water above the junction of the Mohawk; 
Lansingburgh was furnished with water from the hills east of 
the town, and the village of Green Island obtained its water 
from wells and a filter-gallery driven into sandy soil. The others 
used Mohawk or Mohawk-Hudson water.* 
The several intakes are indicated on the chart by squares. 
Under date of April n, 1891, the Health Officer of Schenec- 
tady wrote the author: " The marked increase in typhoid 
fever began in July, 1890, and has just let up. We have had 
about 300 cases. Doctors have not been particular in report- 
ing them, and we have had so many cases of anomalous fevers 
that diagnosis is questionable. Seventy deaths have been re- 
ported." It was not the rule during this epidemic for physi- 
cians to do their whole duty in reporting cases. In one instance 
only two or three cases were reported out of twenty-five. 
Schenectady is a very old town (at that time of 20,000 inhabi- 
* Albany, Cohoes and Waterford have introduced modern filtration plants, 
Troy has abandoned the Hudson River as a source of supply, and West Troy 
(now Watervliet) is building works designed to furnish filtered water from the 
hills. 28 
WATER-SUPPLY 
tants), and its sewerage system delivered the sewage without 
treatment to the Mohawk river. 
The following information was obtained from Dr. Leo, 
Health Officer of Cohoes, and from Dr. Peltier, his predeces- 
sor. Population of Cohoes was 22,000. 
" The epidemic of typhoid began in Cohoes about the end of 
October, 1890, and ceased about the middle of the following 
Cchenectaoy to Cohoes. /7 M/les, 
Cohoes » WTroy. <3 „ 
IV Troy " Cl bany. 6 » 
March. Altogether there were about 1000 cases. The cases 
were mild in character, resulting in very few deaths. Cohoes 
takes its entire water-supply from the Mohawk and returns its 
sewage into the same river. Boiling of water for drinking 
purposes was recommended, and no typhoid developed among 
families who followed the recommendation, except in those 
instances where members of such families drank unboiled water 
while at work away from home. A portion of the city is DRINKING-WATER AND DISEASE 
29 
owned by the great Harmony Knitting Mills, and is built up 
with tenements for their employees, of which there are many 
hundreds. These tenements are kept in excellent repair and 
the plumbing is the best in the city, but extends to kitchen 
sinks and drains only. No water-closets are employed, as 
each house is furnished with privy vault in back yard. Typhoid 
was especially bad in this quarter." The Cohoes Health Officer 
had professional occasion to visit in Waterford (population 
5400) and in Lansingburgh (population 10,000), which towns 
are connected with Cohoes by bridges. He reported that 
hardly a case existed in either of those towns, and it is to be 
noted that one of them took water from the upper Hudson 
above the Mohawk junction, and the other was supplied from 
the hills. 
West Troy (now Watervliet) is situated on the Hudson 
and sewers therein, but by aid of the chart it will be noticed 
that its water-supply came from the Mohawk some distance 
above Cohoes. Its population was 13,000. The following 
information was obtained from Dr. McNaughton, Health 
Officer: 
" Epidemic typhoid began the last of November. At 
meeting of Health Board, about December 15th, fifty cases 
were reported. Of these, forty-two had used Mohawk water, 
the remainder well-water. On December 20th, the Mohawk 
supply was shut off and arrangements made with the town 
of Green Island (which village, by the way, had no fever) 
to use a portion of its supply. One week thereafter the weekly 
report of cases showed fifteen, and the second week there- 
after but one case was reported. The Green Island supply 
was used one month. Upon returning to the Mohawk supply 
in the middle of January, a slight increase in typhoid was 
observed. Total number of cases exceeded 100. The fever, 
as in other places, was very mild, resulting in ten deaths." 
Troy is situated directly opposite West Troy. Its popula- 
tion at that time was 65,000. Its water-supply came partly 
from lakes back in the hills, and partly from the Hudson above 
the Mohawk junction. 30 
WATER-SUPPLY 
There were very few cases of typhoid in Troy during the 
year, and of those few a large percentage were imported from 
Schenectady and West Troy. 
Troy dumped 8,000,000 gallons of sewage into the Hudson 
daily. 
Six miles below Troy is the city of Albany; population, 
100,000. Albany, at the time of the epidemic, took its water 
through an intake in the side of the wharf, directly in front 
of the city. Not only did sewage from the upper Hudson 
and Mohawk flow toward it, but during flood tide and south 
wind the return of its own sewage from the sewer outfalls below 
had been proven by the floating of buoys. 
The typhoid epidemic began in Albany the last of Decem- 
ber, 1890. The disease was mild in character, resulting in 
sixty-two deaths during the months of January, February, 
and March, 1891. The total number of typhoid cases reported 
during the same period was 411, but this figure was known to 
have been unreliable. Albany experienced a very serious epi- 
demic during the winter of 1890-91, and the alarm was wide- 
spread, of that there can be no question. A small portion of 
the city receives its water-supply from an inland gravity 
source. Typhoid was not nearly so plenty in that section, only 
eighteen cases having been reported to the end of March. 
At Van Wie's point, four miles below Albany, the laborers 
employed in cutting ice for the great ice-houses had typhoid 
fever break out among them during January. They used 
ri ver-water for drinking purposes. 
A more recent typhoid epidemic occurred at Albany follow- 
ing the great flood in the Hudson River which submerged and 
threw out of commission the municipal filter, thereby filling 
the entire water-supply system with raw river water. 
The river rose very rapidly to an unprecedented height on 
March 26th, 27th and 28th, 1913, overtopping the filtration 
plant. Throughout the middle of the following month the city 
was visited by a pronounced outbreak of typhoid fever which 
began about April 10th, reached its climax on the 16th and then 
diminished as rapidly as it rose. In all there were about 180 DRINKING-WATER AND DISEASE 
31 
Gauge ReadlngX Hudson River 
>-N»L i_ to fc 
Flltei • Plant Inundated 
Public warned to boil yvater. 
Prospect Reservoir Sterilized, 1st Treatment. 
Prospect:lesenoir Sterilized, 2nd Treatment. 
Date of Last Trt ce of Contaminat ion of Supp y as 
siiown by Laboratory Analysis. 
Public Notified Vlater was Sa 'e. 
I I I I I I 
STATE DEPARTMENT OF HEALTH 
CHART SHOWING 
PREVALENCE OF TYPHOID FEVER 
IN ALBANY FOLLOWING CONTAMINATION 
_ OF WATER SUPPLY BY FLOODS OF 
MARCH 26 - APRIL 2, 1913. 
Theodore Horton. 
n CHIEF ENGINEER. 
2 Cases I eported (Is ; Reported) 
9 Additic nal Cases Reported. 
22 " a " 32 
WATER-SUPPLY 
cases. The average period of incubation may be placed at 
sixteen days. 
Immediately upon the flooding of the filter the people were 
notified to boil their drinking water and arrangements were 
also made to disinfect the reservoirs upon the hills with 
" chloride of lime." This disinfection was accomplished very 
promptly, even before the early cases of the fever began to 
appear, but of course, it could not protect against the infec- 
tion of the first few days. 
Had the warning to 11 boil the drinking water " been heeded 
the epidemic might have been avoided, but the public is always 
slow to take such precaution. The physicians also were slow 
in recording their cases: " In other words the epidemic had 
largely disappeared before the cases had begun to be reported." * 
This epidemic furnished incidental demonstration of how 
great a service the filter plant is daily doing in protecting 
the people against the danger of upstream sewage pollution. 
An important case of typhoid outbreak is added to those 
given, because, though often quoted, and doubtless familiar 
to most readers, the lesson it teaches is too valuable to risk 
losing it from the memory. The description is in part by Dr. 
E. F. Smith: 
The city of Plymouth, Pa., contained in 1885 a population 
of about 8000. In this small community within a period of a 
few weeks there were 1104 cases and 114 deaths from typhoid 
fever. The epidemic was studied on the spot by competent 
New York and Pennsylvania physicians, so that no doubt is 
left either as to the nature of the disease nor as to the method 
of its introduction and spread.f The facts, sifted and tested 
by rigid and expert scientific methods, are as follows: The 
general water-supply of Plymouth is obtained from a moun- 
tain brook, a number of dams being thrown across the stream 
for the purpose of securing reservoir storage, as shown by the 
* Bulletin N. Y. State Dept. Health, May, 1913. 
f " Report upon the Epidemic of Typhoid Fever at Plymouth, Pa." By 
Lewis H. Taylor, M.D., of Wilkesbarre, Pa. DRINKING-WATER AND DISEASE 
33 
following map. During part of the winter of 1884-5, owing 
to the deep freezing of this stream, the hydrant water was 
taken from the neighboring river, Susquehanna. There are 
very few houses on the banks of the mountain brook, and 
Scale 
rinch=2300 feet 
Reservoir No. 4^ 
5,000,000 gals. B 
/ House whence, 
V pollution started. 
Reservoir No. 3 
3,000,000 gals. 
The water is piped 
from No. 1 to the town. 
Reservoir No. 2 
1,700,000 gals. 
Reservoir No. 1C 
300,000 gals. 
TOWN OF 
PLYMOUTH 
SKETCH MAP ILLUSTRATING THE TYPHOID FEVER EPIDEMIC AT PLYMOUTH, PA. 
Reduced from report of Pennsylvania State Board of Health, 1885. 
it would seem that the stream is unusually well protected 
from sources of contamination. On January 2, 1885, dur- 
ing the time that the stream was frozen, a man came from 
Philadelphia ill of typhoid fever. He had contracted the 
disease at a place from which three other persons, ill of fever, 
had been removed to the hospital. This man was cared for 34 
WATER-SUPPLY 
in a house near the stream and just below Reservoir No. 4. 
The discharges from the bowels of the ill man were not 
disinfected. They were thrown by the nurse on the deep 
snow of a side hill sloping toward the stream which was not 
over forty feet distant. A sudden rise in the temperature 
on the 26th of March caused a general thaw, and the melted 
snow of the hillside with its mass of typhoid dejecta, the 
accumulation of nearly three months, was swept into the stream. 
At just this time, owing to the rise of the water in the brook 
and reservoir, the Susquehanna river-water was shut off from 
the water-mains, and that of the brook turned on again. In 
this way the polluted water was pumped to all parts of the 
city. In about two or three weeks hundreds of cases of fever 
developed, and these were confined exclusively to persons who 
used the hydrant water. No cases were traceable to well- 
water except much later, and by secondary infection. Whole 
groups of families using well-water or river-water exclusively 
escaped entirely. In parts of the city where the use of well- 
water was the rule and the use of hydrant water the exception 
only those families suffered which used the latter. One notable 
instance is mentioned by Dr. Taylor where in the upper end of 
the city one family only suffered from the disease. It was 
supposed at first that all in that vicinity used well-water, but 
further inquiry showed this one family to have been in the 
habit of catching and using the hydrant water which leaked 
from the main aqueduct on its way down into the city, prefer- 
ring the supposedly pure water of the mountain stream to that 
of the foul wells of the neighborhood. 
" The epidemic continued with diminished virulence on 
through the Spring and Summer. 
" The chief peculiarity of the epidemic was the gravity of 
the earlier cases. As one medical man expressed it, the earlier 
patients seemed brimful of the poison. From the 10th to the 
20th of April nearly every patient had violent delirium and 
a very high temperature; after that the epidemic seemed to 
settle down into more regular typhoid fever." * 
* Pa. State Board of Health Report, 1885. DRINKING-WATER AND DISEASE 
35 
This change in virulence is most instructive and points 
at once to the concentrated doses of fresh infective material 
delivered to the early cases. 
Sundry data of the epidemic: 
Cases in April 713 . Cases in May... . 261 
Cases in June 83 Cases in July 31 
Cases in August. . . 15 Cases in September 1 
1104 
Average age of those who died 24 years 
Average length of illness of those who died.. . 23 days 
Average length of illness of those recovered. . 58 days 
Fatality rate. io|% 
The cost of this outbreak, in actual cash, is fortunately 
well established, and is deserving of attention by those 
charged with the care of the public health. It is itemized as 
follows: 
Loss of wages for those who recovered.$3o,o2o.o8 
Care of the sick 67,100.17 
Yearly earnings of those who died. ... 18,419.52 
It should be noted here that a " spot-map " of an epidemic 
of " water-borne typhoid " shows the 11 spots " distributed 
evenly over the city. This would be expected as the water 
bearing the disease is delivered to all the people. (See next page.) 
Very similar to the Plymouth outbreak was that at New 
Haven during the first part of 1901. Up to April there had 
been 315 cases reported. 
" In the remote limits of an adjoining town, eight miles 
from New Haven, there had been during the first three months 
of the year, at different times, three cases of typhoid. The 
discharges from the patients were not disinfected. 
" During the illness of these patients the ground was frozen 
solid, and there is every reason to believe that there was no 
true burial of the feces. The alleged place of burial was on a 36 
WATER-SUPPLY 
side hill with a steep inclination to a watercourse leading 
directly to Dawson Lake, an important reservoir of the New 
Haven Water Company." 
When it is remembered that much of the Chicago sewage 
flowed into Lake Michigan, and that until recently the intakes 
PRESQUE ISLE BAY 
LAKE 
ERIE 
SPOT -MAP SHOWING EVEN DISTRIBUTION OF CASES DURING TYPHOID FEVER 
EPIDEMIC, 1911. 
(Report of Com. Health, Pa., 19n.) 
supplying the city with the lake-water were situated only a 
few hundred feet off shore, a comparison of the typhoid death- 
rates per 100,000 inhabitants for periods of five years before 
and after the driving of the four-mile tunnel is suggestive.* 
The tunnel was opened December 3, 1892. 
Average for the years 1888 to 1892 inclusive 96.9 
11 1893 to 1897 " 42.1 
* Bui. Chicago Health Depart., May 27, 1905. DRINKING-WATER AND DISEASE 
37 
Much as the new tunnel did for Chicago, the difficulty 
was not entirely remedied, for the reason that sewage occa- 
sionally reached the four-mile crib until after the opening of the 
Drainage Canal, which caused it to flow into the Illinois River. 
As a measure of the betterment in the quality of the lake 
water resulting from the opening of the canal, let the following 
facts, taken from the Bulletin of the Department of Health of 
Chicago for May 19, 1906, speak for themselves: ''During 
the six years immediately preceding the opening of the drainage 
channel there had been a total of 141,473 deaths in Chicago, 
an average of 23.579 each year, a rate of 16.2 per thousand 
of the population. During the subsequent six years, ending 
December 31, 1905, there was an annual average of 26,373 
deaths, a rate of 14.31 per thousand, or a reduction of 11.7 
per cent in the general mortality. 
" Deaths from typhoid fever in the pre-channel period were 
3275, an annual rate of 37.6 per 100,000 population. Deaths 
from typhoid fever during the channel period were 2937, an 
annual rate of 26.9 per 100,000 population, a reduction of 29 
per cent. 
" Deaths from diarrhoeal diseases in the pre-channel period 
16,669, an annual rate of 192.9 per 100,000 population. Deaths 
from diarrhoeal diseases during the channel period 13,609, an 
annual rate of 123.1 per 100,000 population, a reduction of 
36.9 per cent. 
"The total cost of the channel up to December 31, 1905, 
had been less than 50 million dollars. 
" The figures given above of typhoid-fever and diarrhoeal-dis- 
eases mortality show a constructive or potential saving of 
8966 lives, which, at the legislative valuation, represents a 
total of $89,660,000." 
It will be noted from the foregoing that typhoid fever is not 
the only disease that is diminished by an improved water supply 
but that the general group of intestinal disorders responds 
as well. This point is well illustrated in a paper by M. N. 
Baker, read before the New England Water Works Association.* 
* J. N. E. Water Works Asso., xx. 163. 38 
WATER-SUPPLY 
Whipple also enlarges upon the same proposition, as will be 
seen from the following table where comparison is made be- 
tween the rates in Albany, both before and after filtration, 
and those in Troy, where no filtration is used, during the same 
periods.* 
ALBANY (BEFORE AND AFTER FILTRATION) 
Average Annual Rates per 
100,000 Population. 
1894-98 
1900-04 
Differ- 
ence. 
Reduc- 
tion as 
Per Cent. 
Typhoid fever ■ • 
104 
26 
78 
75 
Diarrhoeal diseases 
12 5 
^7 
Children under five years. 
606 
00 
TOO 
297 
o / 
4Q 
Total deaths 
2246 
1868 
378 
17 
TROY (SAME PERIOD, NO FILTRATION) 
Average Annual Rates per 
100,000 Population. 
1894-98 
1900-04 
Differ- 
ence. 
Reduc- 
tion as 
Per Cent. 
Typhoid fever 
57 
57 
0 
O 
Diarrhoeal diseases 
116 
102 
14 
12 
Children under five years 
531 
435 
96 
18 
Total deaths 
2157 
2028 
129 
6 
TABLE SHOWING IMPROVEMENT IN DEATH-RATES FROM TYPHOID 
FEVER IN MASSACHUSETTS 
Deaths per 
Years. 
100,000 
Inhabitants. 
1871-75 
82 
1876-80 
42 
1881-85 
41 
1886-9O 
46 
1891-95 
34 
1896-I9OO 
26 
I9OI-O5 
T9 
I9O6-IO 
13-8 
7-6 
* J. New England Water Works Asso., xix. 461. DRINKING-WATER AND DISEASE 
39 
TYPHOID FEVER DEATH-RATES PER 100,000 POPULATION, IN CITIES OF THE UNITED STATES OF OVER 100,000 POPULATION* 
City. 
Pop ula- 
tion 1910 
Average 
for 1900- 
J1905, incl. 
1906 
1907 
1908 
1909 
1910 
1911 
1912 
Source of Water-supply. 
Purification of Water-supply.J 
Allegheny, Pa.f 
150,000 
107 
136 
97 
40 
28 
42 
39 
9 
Allegheny River 
Sterilization 
Albany -N. Y 
100 253 
25 
20 
20 
11 
19 
15 
18 
18 
Hudson River 
Slow sand filtration 
Atlanta, Ga 
154,839 
65 
50 
64 
47 
44 
43 
56 
35 
Chattahooche River 
Rapid sand filtration 
Baltimore, Md.... ... 
558,845 
36 
34 
41 
31 
23 
42 
26 
22 
Gunpowder Cr. and Jones Falls 
Sterilization 
Birmingham, Ala 
132,685 
50 
39 
71 
64 
44 
51 
44 
38 
Cahaba River.. . 
Rapid sand filtration 
Boston, Mass 
680,585 
23 
22 
10 
26 
14 
11 
8 
9 
Impounded streams 
None 
Bridgeport, Conn 
102,054 
15 
10 
13 
13 
13 
9 
4 
8 
Poland River Impounded 
None 
Buffalo, N. Y 
423,715 
29 
24 
29 
21 
23 
20 
25 
11 
Lake Erle 
None 
Cambridge, Mass 
104,839 
18 
18 
10 
10 
9 
12 
4 
4 
Impounded streams 
None 
Chicago, Ill 
2,185,283 
27 
18 
18 
15 
12 
14 
10 
7 
Lake Michigan 
One-fourth sterilized 
Cincinnati, O 
363,591 
54 
71 
46 
19 
13 
6 
11 
7 
Ohio River 
Rapid sand filtration 
Cleveland, O 
560,663 
51 
20 
19 
13 
12 
19 
14 
6 
Lake Erie 
Sterilization 
Columbus, O 
181,511 
61 
45 
38 
110 
17 
17 
14 
19 
Scioto River 
Rapid sand filtration 
Dayton, O 
116,577 
29 
28 
38 
16 
24 
18 
18 
17 
Driven wells 
None 
Denver, Colo 
213,381 
37 
68 
67 
58 
24 
28 
17 
13 
Platte River and Marston Lake 
Infiltration galleries, slow and 
Detroit, Mich 
465,766 
17 
22 
28 
22 
19 
20 
16 
19 
Lake St. Clair 
None rapid sand filters 
Fall River, Mass 
119,295 
19 
8 
18 
13 
14 
13 
15 
16 
No. Wattupa Pond 
None 
Grand Rapids, Mich... 
112,571 
34 
39 
30 
30 
17 
26 
25 
31 
Grand River 
Rapid sand filtration 
Indianapolis, Ind 
233,650 
76 
39 
29 
26 
22 
30 
26 
17 
White River and wells 
Slow sand filtration 
Jersey City, N. J 
267,779 
19 
20 
14 
10 
8 
12 
8 
8 
Rockaway River Impounded 
Sterilization 
Kansas City, Mo 
248,381 
48 
38 
40 
35 
23 
43 
24 
12 
Missouri River 
Sterilization 
Los Angeles. Cal 
319,198 
35 
18 
23 
19 
18 
12 
13 
12 
Los Angeles River and wells 
Infiltration galleries 
Louisville, Ky 
223,928 
55 
63 
79 
49 
43 
31 
25 
19 
Ohio River 
Rapid sand filtration 
Lowell, Mass 
106,294 
19 
7 
9 
24 
11 
21 
6 
9 
Driven wells 
None 
Memphis, Tenn- 
131,105 
37 
39 
35 
33 
41 
28 
61 
56 
Wells 
None 
Milwaukee, Wls 
373,859 
19 
31 
26 
17 
21 
46 
19 
23 
Lake Michigan 
Sterilization 
Minneapolis, Minn. . . 
301,408 
38 
33 
26 
18 
20 
58 
11 
11 
Mississippi River 
Sterilization 
Nashville, Tenn 
110,364 
54 
66 
62 
53 
48 
51 
29 
Cumberland River 
Coagulation-sedimentation 
Newark, N. J 
347,469 
17 
18 
24 
12 
11 
13 
10 
7 
Pequannock River Impounded 
None 
New Haven, Conn.. . . 
133,605 
44 
54 
30 
34 
20 
18 
23 
24 
Lakes 
Slow sand filtration (Lake Whit- 
New Orleans, La 
339,075 
40 
30 
56 
31 
25 
32 
30 
13 
Mississippi River 
Rapid sand filtration ing only) 
New York, N. Y 
4,766,883 
19 
15 
17 
12 
12 
12 
11 
10 
Croton River Impounded 
Sterilization 
Oakland, Cal 
150,174 
19 
26 
28 
18 
7 
13 
13 
19 
Impounded 
Rapid sand filtration 
Omaha, Neb 
124,096 
20 
28 
24 
22 
31 
89 
17 
12 
"Missouri River 
Sedimentation and sterilization 
Paterson, N. J 
125,600 
25 
4 
11 
10 
5 
7 
5 
7 
Passaic River 
Rapid sand filtration 
Philadelphia, Pa 
1,549,008 
47 
74 
60 
36 
22 
17 
14 
12 
Delaware and Schuylkill Rivers 
Slow sand filtration 
Pittsburgh, Pa.f 
410,000 
132 
141 
135 
53 
13 
12 
10 
6 
Allegheny River 
Slow sand filtration 
Portland, Ore 
207,214 
25 
33 
24 
20 
23 
18 
17 
Bull Run River 
None 
Providence, R. I 
224,326 
20 
19 
8 
16 
12 
17 
14 
10 
Pawtucket River 
Slow sand filtration 
Richmond, Va 
127,628 
66 
44 
41 
50 
24 
22 
17 
15 
James River 
Coagulation-sedimentation 
Rochester, N. Y 
218,149 
15 
17 
16 
12 
9 
14 
11 
12 
Hemlock Lake 
None 
St. Louis, Mo 
687,029 
33 
18 
16 
15 
13 
10 
13 
Mississippi River 
Coagulation-sedimentation 
St. Paul, Minn 
214,744 
14 
21 
17 
12 
20 
20 
10 
9 
Lake and artesian wells 
None 
San Francisco, Calif. . 
416,912 
20 
57 
27 
17 
15 
16 
14 
Impounded 
Filtration-aeration 
Scranton, Pa 
129,867 
18 
ii 
76 
11 
11 
14 
12 
9 
Impounded 
Rapid sand filtration 
Seattle, Wash 
237,194 
20 
48 
22 
23 
14 
10 
8 
Cedar Lake and River 
None 
Spokane, Wash 
104,402 
47 
38 
40 
32 
32 
37 
17 
Dug wells 
None 
Syracuse, N. Y 
137,249 
i4 
10 
16 
15 
12 
28 
15 
16 
Skaneateles Lake 
None 
Toledo, O 
168,497 
36 
45 
36 
40 
31 
37 
22 
31 
Maumee River 
Rapid sand filtration 
Washington, D. C. .. . 
331,069 
59 
52 
36 
39 
33 
25 
22 
23 
Potomac River 
Slow sand filtration 
Worcester, Mass 
145,986 
17 
12 
14 
10 
8 
16 
7 
3 
Brooks and pond 
None 
* Compiled by Geo. A. Johnson. + Allegheny and Pittsburgh are now combined in 
Greater Pittsburgh. t The rapid sand nitration plant at Atlanta, Ga., is of 
the pressure type; all 
the other rapid Alters are of the gravity type. 40 
WATER-SUPPLY 
HEALTH STATUS OF THE WORLD'S LEADING CITIES FOR THE YEAR 
1912 * 
City. 
' Population. 
General Death 
Rate per 1000 
of Pop. 
Typhoid Fever 
Death-Rates 
per 100,000 
of Pop. 
London (proper) 
4,519,754 
13.6 
3 
New York 
5,173,064 
14.1 
IO 
Paris 
2,872,128 
16.3 
9 
Chicago 
2,294,711 
. 14.8 
8 
Berlin 
2,083,391 
14-4 
2 
Vienna 
2,081,335 
15-4 
2 
St. Petersburg 
1,664,000 
21.9 
50 
Moscow 
1,640,900 
24-3 
12 
Philadelphia '... 
1,606,105 
T5 i , 
12 
Buenos Aires 
Bombay 
979,445 
39-8 
Hambuig 
975,562 
136 
3 
Rio de Janeiro 
946,134 
21.3 
6 
Budapest 
910,548 
18.5 
II 
Calcutta 
896,067 
28.1 
Brussels 
745,38o 
13-5 
7 
Boston 
8 
Cairo . . 
704,956 
37-4 
Sydney 
674,500 
11.4 
9 
Milan 
6i5,375 
15.8 
Munich 
615,000 
14-7 
I 
Melbourne 
614,300 
14.0 
5 
Leipzig 
605,755 
12.7 
I 
Amsterdam 
584,416 
II . 2 
3 
Rome.. 
572,099 
14.0 
Baltimore 
569,561 
18.3 
Dresden : 
559,700 
13 I 
3 
Cologne 
532,ooo 
14-4 
2 
Breslau 
53o,ooo 
18.4 
2 
Madras 
518,660 
38.7 
In the State of Connecticut the typhoid statistics for the 
past forty-three years show a continual improvement, which 
must be due, at least in part, to abolition of old private wells 
for new and better water-supply. The number of deaths 
(for the entire State) from typhoid per 100,000 population 
stands as follows:! 
* Bulletin Chicago School of Sanitary Instruction, 
t See Board of Health Reports. DRINKING-WATER AND DISEASE 
41 
Average for the five years 1865-69 83 
" " " 1870-74 75 
" " 11 1875-79 54 
" " 11 1880-84 42 
" 11 11 1885-89 34 
11 " " 1890-94 35 
" " 1895-99 22 
1900-04 24 
1905-09 19 
" three " 1910-12 14 
Statistics compiled for the Massachusetts Board of Health 
by Mr. H. F. Mills * show how greatly the typhoid death-rate 
is improved in towns by a change from the domestic well 
system to that of a public supply. 
CHANGE IN THE DEATH-RATE FROM TYPHOID FEVER PER 100,000 
INHABITANTS IN CITIES OF MASSACHUSETTS WHICH INTRO- 
DUCED WATER-SUPPLIES FROM 1867 TO 1876 
Annual 
Average, 
1859-68. 
Date of 
Water-supply. 
Annual Average. 
1878-89. 
Per Cent of 
Average. 
Fall River 
77-8 
1874 
63.2 
81 
Springfield 
96.7 
1875 
52.9 
55 
Taunton 
61.2 
1876 
50-2 
82 
Northampton 
109.8 
1871 
40-4 
37 
Lynn 
90.6 
1871 
38.7 
43 
New Bedford 
77-7 
1869 
38.0 
49 
Newton 
65-7 
1876 
36.5 
56 
Malden 
80.4 
1870 
35-4 
44 
Fitchburg 
i°5-9 
1872 
31.6 
3° 
Woburn 
82.9 
1873 
29-5 
36 
Somerville 
42.8 
1867 
29-5 
69 
Chelsea 
59-7 
1867 
28.9 
48 
Waltham 
81.2 
1873 
24.2 
30 
Average 
79-4 
38.3 
The advance along sanitary lines shown by the . foregoing 
data is encouraging, and it effectually answers the frequently 
recurring question, " How was it our fathers got along so well 
without all these so-called modern improvements? " 
* Mass. Board of Health, 1890. 42 
WATER-SUPPLY 
Figures such as those given, and others like them that might 
be quoted, stand in evidence that the question is a begged one, and 
that our progenitors were not so well off as many people fancy. 
In illustration of this point, it is instructive to note the 
improvement in the general death-rate of London. 
DEATHS IN LONDON FROM ALL CAUSES PER XOOO POPULATION PER ANNUM IN 
PERIODS REPRESENTING THE 17TH, 18TH, AND 19TH CENTURIES. 
(From report of State Board of Health, Michigan.) 
Improvement in the general death-rate per year for New York 
City per 1000 population is shown by the following: 
Death-rate per 1000 
1872 . 3O-96 
1882 28.06 
1892 25.38 
1902 18.58 
1912 14.II 43 
DRINKING-WATER AND DISEASE 
There is no question but that sanitary betterments show 
widespread effect upon the general death-rates. During the 
DEATHS 
IN 1000 DEATHS 
„ FROM ALL CAUSES ' 
6 r- 
DEATHS 
IN 1OOO DEATHS 
FROM ALL CAUSES ' 
-1-1 I I I I. ..I I 
OPENING OF WATERWORKS 
COMMEh CEMENT OF SEWERAGE\oPERA~TQN 5. 
TYPHOID MORTALITY IN BERLIN (rOECHLING). 
past thirty years the improvement for ten of the world's largest 
cities averages 31.1 per cent.* 
* See U. S. Census Bui., 109. 44 
WATER-SUPPLY 
In this connection it is interesting to observe the increase 
in the median age of the population of the United States.* 
1850 18.28 years 
i860 . 18.87 " 
187° z9-65 " 
1880 20.45 " 
1890 21.38 " 
1900 22.3 (estimated) 
It might be well to note that water-supply systems should 
not be introduced into country towns until arrangements 
have been made to care for the necessarily increased volume 
of sewage thereby produced; otherwise the old vaults may 
become overloaded and carry pollution to greater distances 
than it went before, thus damaging the remaining wells. 
As showing the exceeding difference between care and no 
care in the selection of water for potable supply, the following 
quotation is taken from a report by Dr. Simmons, of the 
Yokohama Board of Health, covering certain features of the 
water question in India: 
" The drinking-water supply is derived from wells, so- 
called ' tanks ' or artificial ponds, and the watercourses of 
the country. The wells generally resemble those in other 
parts of Asia. The tanks are excavations made for the pur- 
pose of collecting the surface-water during the rainy season 
and storing it up for the dry. Necessarily they are mere 
stagnant pools. The water is used not only to quench thirst, 
but is said to be drunk as a sacred duty. At the same time, 
the reservoir serves as a large washing-tub for clothes, no 
matter how dirty or in what soiled condition, and for personal 
bathing. Many of the watercourses are sacred: notably the 
Ganges, a river 1600 miles long, in whose waters it is the religious 
duty for millions, not only for those living near its banks, 
but of pilgrims, to bathe and to cast their dead. The Hindoo 
cannot be made to use a latrine. In the cities he digs a hole 
* Merriman's " Sanitary Engineering " (1899), page 43. DRINKING-WATER AND DISEASE 
45 
in his habitation; in the country he seeks the fields, the hill- 
sides, the banks of streams and rivers when obliged to obey 
the calls of nature. Hence it is that the vicinity of towns and 
the banks of the tanks and watercourses are reeking with 
filth of the worst description, which is of necessity washed 
into the public water-supply with every rainfall. Add to 
this the misery of pilgrims, their poverty and disease, and their 
terrible crowding into the numerous towns which contain some 
temple or shrine, the object of their devotion, and we can see 
how India has become and remains the hot-bed of the cholera 
epidemic. In the United States official report, the horrors 
incident upon the pilgrimages are detailed with appalling 
minuteness. W. W. Hunter, in his Orissa, states that twenty- 
four high festivals take place annually at Juggernaut. At one 
of them, about Easter, 40,000 persons indulge in hemp and 
hasheesh to a shocking degree. For weeks before the car 
festival in June and July, pilgrims come trooping in by thou- 
sands every day. They are fed by the temple cooks to the 
number of 90,000. Over 100,000 men and women, many of 
them unaccustomed to work or exposure, tug and strain at 
the car until they drop exhausted and block the road with 
their bodies. During every month of the year a stream of 
devotees flows along the great Orissa road from Calcutta, and 
every village for three hundred miles has its pilgrim encamp- 
ments. The people travel in small bands, which at the time 
of the great feasts actually touch each other. Five-sixths of 
the whole are females, and ninety-five per cent travel on foot, 
many of them marching hundreds and even thousands of 
miles, a contingent having been drummed up from every 
town or village in India by one or other of the three thousand 
emissaries of the temple, who scour the country in all direc- 
tions in search of dupes. When those pilgrims who have not 
died on the road arrive at their journey's end, emaciated, with 
feet bound up in rags and plastered with mud and dirt, they 
rush into the sacred tanks or the sea, and emerge to dress in 
clean garments. Disease and death make havoc with them 
during their stay; corpses are buried in holes scooped in the 46 
WATER-SUPPLY 
sand, and the hillocks are covered with bones and skulls washed 
from their shallow graves by the tropical rains. The temple 
kitchen has the monopoly of cooking for the multitude, and 
provides food which, if fresh, is not unwholesome. Unhappily, 
it is presented before Juggernaut, so becoming too sacred for 
the minutest portion to be thrown away. Under the influence 
of the heat it soon undergoes putrefactive fermentation, and in 
forty-eight hours much of it is a loathsome mass, unfit for 
human food. Yet it forms the chief sustenance of the pilgrim, 
and is the sole nourishment of thousands of beggars. Some 
one eats it to the very last grain. Injurious to the robust, it 
is deadly to the weak and wayworn, at least half of whom 
reach the place of suffering under some form of bowel com- 
plaint. Badly as they are fed, the poor wretches are worse 
lodged. Those who have the temporary shelter of four walls 
are housed in hovels built upon mud platforms about four feet 
high, in the centre of each of which is the hole which receives 
the ordure of the household, and around which the inmates 
eat and sleep. The platforms are covered with small cells 
without any windows or other apertures for ventilation, and in 
these caves the pilgrims are packed, in a country where, during 
seven months out of the twelve, the thermometer marks from 
85° to ioo° Fahr. Hunter says that the scenes of agony and 
suffocation enacted in these hideous dens baffle description. 
In some of the best of them, 13 feet long by 10 feet broad and 
6| high, as many as 80 persons pass the night. It is not, 
then, surprising to learn that the stench is overpowering and 
the heat like that of an oven. Of 300,000 who visit Jugger- 
naut in one season, 90,000 are often packed together for a 
week in 5000 of these lodgings. In certain seasons, however, 
the devotees can and do sleep in the open air, camping out in 
regiments and battalions, covered only with the same meagre 
cotton garment that clothes them by day. The heavy dews 
are unhealthy enough; but the great festival falls at the begin- 
ning of the rains, when the water tumbles in solid sheets. 
Then lanes and alleys are converted into torrents or stinking 
canals, and the pilgrims are driven into the vile tenements. DRINKING-WATER AND DISEASE 
47 
Cholera invariably breaks out. Living and dead are huddled 
together. In the numerous so-called corpse-fields around the 
town as many as forty or fifty bodies are seen at a time, and 
vultures sit and dogs lounge lazily about gorged with human 
flesh. In fact, there is no end to the recurrence of incidents 
of misery and humiliation, the horrors of which, says the 
Bishop of Calcutta, are unutterable, but which are eclipsed 
by those of the return journey. Plundered by priests, fleeced 
by landlords, the surviving victims reel homeward, staggering 
under their burden of putrid food wrapped up in dirty clothes, 
or packed in heavy baskets or earthenware jars. Every 
stream is flooded, and the travellers have often to sit for days 
in the rain on the bank of a river before a boat will venture to 
cross. At all these points the corpses lie thickly strewn around 
(an English traveller counted forty close to one ferry), which 
accounts for the prevalence of cholera on the banks of brooks, 
streams, and rivers. Some poor creatures drop and die by 
the way; others crowd into the villages and halting-places 
on the road, where those who gain admittance cram the lodg- 
ing-places to overflowing, and thousands pass the night in the 
streets, and find no cover from the drenching storms. Groups 
are huddled under the trees; long lines are stretched among 
the carts and bullocks on the roadside, their hair saturated 
with the mud on which they lie; hundreds sit on the wet 
grass, not daring to lie down, and rocking themselves to a monot- 
onous chant through the long hours of the dreary night. It 
is impossible to compute the slaughter of this one pilgrim- 
age. Bishop Wilson estimates it at not less than 50,000. 
And this description might be used for all the great 'Indian 
pilgrimages, of which there are probably a dozen annually, 
to say nothing of the hundreds of smaller shrines scattered 
through the peninsula, each of which attracts its minor hordes 
of credulous votaries. So that cholera has abundant oppor- 
tunities for spreading over the whole of Hindostan every 
year by many huge armies of filthy pilgrims; and the country 
itself well deserves the reputation it universally possesses of 
being the birthplace and settled home of the malady." 48 
WATER-SUPPLY 
With the Chinese it is quite different. " Although their 
country is in closest proximity to India, and of much greater 
extent and twice as populous, you will find that cholera is 
comparatively rare. The drinking-water supply of China is 
derived from wells, springs, and natural streams. Now, 
though the wells and springs are used in China for drinking 
purposes to much the same extent and in much the same 
manner as in India, yet the rivers and lakes are not drunk from 
as a part of a religious duty, nor is bathing in them a sacred 
rite. The absence of pilgrimages contributes to keep the 
water comparatively uncontaminated. 
" Human manure is valuable and hoarded for fertilizing 
purposes. Hence the excreta are deposited by the individual 
in a receptacle made for the express purpose, and from motives 
of economy kept in a fairly good condition of repair. Even 
in cities and large towns latrines are not employed. Special 
wooden boxes are among the first necessities of bedroom 
furniture, and form part of every bridal outfit. The contents 
are daily emptied into earthen jars or wooden tubs placed in 
the court-yard of the house, whence they are removed by the 
scavenger either direct to the fields, or to boats destined to 
convey them to a distance. Thus the greatest amount of 
security attainable is provided against the contamination of 
the water-supply from this source. A still more potent pre- 
ventive of infection is to be found in the fact that the Chinese 
will always, if possible, boil water before drinking it, even if 
they are unable to make it into some kind of tea. Here it is 
easy to see, in the contrast between the customs of the Hindoo 
on the one hand and the Chinese on the other, how in the one 
case every possible facility is provided for the propagation of 
infection; in the other, how the danger of contamination is 
reduced to a minimum." 
This statement concerning China is hardly in accord with 
the following naval report: 
11 In Japan and China the close relation of the food and 
water supply with the excreta not only illustrates the aetiology 
of cholera, but, at the same time, shows what small prospect DRINKING-WATER AND DISEASE 
49 
there is of its extermination. In Japan the soil is tilled in 
absolute contiguity to the wells, and is fertilized with liquefied 
human excreta. Dr. Jameson, a physician of Shanghai, cites 
an instance where, under a spigot, he saw the rice for the 
daily food being washed at the same time with a vessel just 
emptied of cholera discharges." * 
Illustrations such as have been given, showing the power 
of water to carry specific disease, could be very greatly multi- 
plied. J. W. Hill has published a list of cases of this character 
in " J. Am. Water-works Asso.," 1897, 168; Wilson's "Hand- 
book of Hygiene " furnishes a further number, notably the 
celebrated " Broad Street well case "; the widely known 
report of the epidemic at Lausen, Switzerland, will be found 
in " Nature," xiii, 447; while a historic summary of two hundred 
and five outbreaks of " water-borne typhoid " in Great Britain 
has been published by E. Hart, in a report prepared for the 
Parliamentary Bills Committee of the British Medical Asso- 
ciation, 1897. It would not be wise to devote space to these 
cases beyond the reference given, but it is well to pause for 
a moment to consider what may be learned from the terrible 
outbreak of cholera at Hamburg, Germany, in 1892. 
The city had at that time a population of 640,400. In the 
official tabulation the epidemic is noted as having lasted from 
August 16th to November 12th, although 42 deaths appear 
to have taken place in the next month (December) and 20 in 
January, 1893. The total number of cholera cases reported 
during that time was 17,020, with a total death-list of 8605, 
a mortality percentage of 50.05. 
By months the cases were: 
December.... 42 
January 20 
February 1 
March 1 
17,020 
August 7 >42 7 
September 9,341 
October 181 
November 7 
* Report of Surgeon-General U. S. Navy, 1892. 50 
WATER-SUPPLY 
To a proper appreciation of the conditions of this epidemic 
a study of the local map is essential. 
It will be observed that Altona (143,000 population), 
Hamburg (640,400 population), and Wandsbeck (20,000 pop- 
A, 3,0, Ou-ifctll.s of mcti'n' fewer#, 
JZa-m b UT^ff-b] 7/toncb-. 
G Jb'taTte' of Ha/nbu rg TWiterlrorfa. 
lumbers -show popitJabo/ty in /887. 
FigG*. JJ J3 larib&ne^e'. Zn^ta Av'a-nxiy 
-a ct/7zd-ZHi&n3/3ilionct^Otiien<s'&7T/yyS^/'l7ybrbt^s. 
AFTER REINKE AND SEDGWICK. 
ulation) are practically one and the same town, separated 
by only imaginary boundaries, which a stranger could not 
locate. The three municipalities are, however, supplied with 
water from three different sources. Wandsbeck obtains filtered 
water from a lake unexposed to contamination; Hamburg DRINKING-WATER AND DISEASE 
51 
pumps water from the Elbe River, and in 1892 the intake was 
situated just south of the city, but not far enough up-stream 
to escape contamination from a recession of polluted water at 
flood-tide. After some imperfect sedimentation, the water 
passed directly to the consumer without filtration. Altona, 
strangely enough, pumps its water from the Elbe at a point 
about eight miles below that at which the river receives the 
combined sewage of the three cities, with their population of 
over 800,000. Fortunately for Altona, this most grossly 
polluted supply is filtered with exceeding care before delivery 
to the people. Further description of the Hamburg epidemic 
can best be given in the words of Dr. Thorne, medical officer 
of the London Local Government Board: * 
" The different behavior of Hamburg and Altona as regards 
cholera is extremely interesting in this connection. The two 
towns adjoin; they are practically one city. The division 
between the two is no more obvious than that between two 
densely peopled London parishes, and yet a spot-map indicat- 
ing the houses which were attacked with cholera, which was 
shown to me by Professor Koch, points out clearly that 
whereas the disease prevailed in epidemic form on the Ham- 
burg side of the boundary line, that line, running in and out 
among the streets and houses and at times passing diagonally 
through the houses themselves, formed the limit beyond which 
the epidemic as such did not extend. The dots on one side 
of the dividing line were proof of the epidemicity of cholera in 
Hamburg; their comparative absence on the Altona side of it 
was proof of the absence of an epidemic in Altona. To use 
Professor Koch's own words: ' Cholera in Hamburg went 
right up to the boundary of Altona and there stopped. In 
one street, which for a long way forms the boundary, there 
was cholera on the Hamburg side, whereas the Altona side 
was free from it.' And yet there was one detectable differ- 
ence, and one only, between the two adjacent areas-they 
had different water-services. 
" Professor Koch has collected certain proofs which he 
* " Cholera Prospects and Prevention," London, 1893. 52 
WATER-SUPPLY 
CHOLERA IN THE TERRITORY ADJACENT TO THE BOUNDARY 
between HAMBURG and ALTONA, 
in the Epidemic of 1892. / 
CHART SHOWING DISTRIBUTION OF CHOLERA CASES IN THE EPIDEMIC OF 1892. 
(From "Filtration of Municipal Water-supply," by Rudolph Hering.) 
Boundary between the two cities. 
• Cases of Cholera, 
o Cases imported from Hamburg. DRINKING-WATER AND DISEASE 
53 
regards as crucial on this point, and Dr. Reincke has supplied 
me with a small plan in support of the contention. At one 
point close to and on the Hamburg side of the boundary line 
between- Hamburg and Altona is a large yard known as the 
Hamburger-Platz. It contains .two rows of large and lofty 
dwellings, containing seventy-two separate tenements and 
some 400 people, belonging almost wholly to those classes 
who suffered most from cholera elsewhere in Hamburg. But 
whilst cholera is shown .by the spot-map to have prevailed all 
around, not a single case occurred amongst the many residents 
of this court during the whole epidemic. And why? Pro- 
fessor Koch explains that owing to local difficulties water 
from the Hamburg mains could not easily be obtained for the 
dwellings in question, and hence a supply had been laid on from 
one of the Altona mains in an adjacent street. This was the 
only part of Hamburg which received Altona water, and I 
am informed that it was the only spot in Hamburg in which 
was aggregated a population of the class in question which 
escaped the cholera. At the date of my visit to Hamburg 
a notice-board was affixed at the entrance of this court. It 
stated that certain tenements were to let; but above all, in 
large type, and as an inducement to intending tenants, was 
the announcement that the court was not only within the 
jurisdiction of Hamburg, with the privileges still attaching 
to the old Hanseatic cities, but that it had a supply of Altona 
water." 
During this epidemic the deaths in the several cities 
were: 
Population. 
Deaths. 
Deaths per 10,000 
Inhabitants. 
Hamburg 
640,400 
8605 
134-4 
Altona 
143,000 
328 
23 0 . 
Wandsbeck 
20,000 
43 
22.0 
" That infectious matter was communicated to the Elbe 
water from Hamburg is not in any way a hypothesis. Cholera 
germs had been as a fact found in the Elbe water. They 54 
WATER-SUPPLY 
were found a little below the place where the Hamburg main 
sewer flows into the Elbe. They were also found in one of the 
two (Altona) basins into which the water flowed before filtra- 
tion." * 
Dr. Reincke says f in regard to Hamburg that the evi- 
dence that the reduction of typhoid in the city has been the 
result of an improvement in the water is substantiated by the 
fact that the typhoid among the shipping interests in the 
harbor, which use the raw water, is as great as it was before, 
while in the city it has fallen from 90 per 100,000 in 1887 to 
6 in 1894 and 9 in 1895. In another place he argues that 
typhoid fever has always come through the same channels as 
cholera, and shows by a tabular statement that the maximum 
of typhoid has followed two or three weeks later than the 
maximum of cholera, the difference in time representing the 
longer incubation period of the former disease. 
No invasion of cholera has yet appeared to test the effi- 
ciency of the elaborate filter-plant which Hamburg has since 
constructed, but the filters are certainly doing good work 
toward the suppression of typhoid fever. 
To the man who purposes dealing with the question of 
public water-supply, some knowledge of the aetiology of the 
two prominent water-borne diseases, " cholera " and " typhoid 
fever," is essential to the proper and successful meeting of 
his professional responsibilities. These two diseases are not 
the only ones that are water-borne, but they are the most 
prominent. Dysentery, paratyphoid, diarrhoea, and sundry 
other disorders are on the list, but if the serious ones be guarded 
against the protective measures taken will serve to prevent 
danger from the others as well. 
The cholera " germ" (spirillum cholera Asiaticce}, or 
11 comma bacillus," was discovered by Koch in 1884 in the 
* Koch, " Zeit. fur Hygiene und Infect.-Krank.," xiv. 
f " Zur Epidemiologie des Typhus in Hamburg und Altona," paper before the 
Hamburg Artz, Verein, Jan. 2, 1896; also Fuertes, " Water and Public Health." DRINKING-WATER AND DISEASE 
55 
excreta of cholera patients and in the intestinal contents of 
those dead of the disease. 
The spirillum will grow in ordinary culture-jelly at the 
usual temperature, forming in twenty-four hours small white 
colonies which increase in size and finally entirely liquefy the 
gelatine. Growth is arrested if the temperature exceed 107° F., 
or if it fall below 590 F. 
In shape it is not unlike the " comma," whence it derives 
its name, and the union of two or more attached end to end 
cholera spirillum. (After Frankel and Pfeiffer.) 
From Hiss and Zinsser's " Bacteriology." Courtesy of D. Appleton & Co.) 
often causes the appearance of semicircles, S-shaped figures, 
and long spiral filaments. In size it varies from 0.8 to 2 
microns in length, and from 0.3 to 0.4 in breadth. It is gen- 
erally conceded that the cholera spirillum does not form spores, 
a characteristic which permits of its ready destruction by heat, 
a " spore " being much more difficult to destroy than a full- 
grown bacterium. Sternberg found the thermal death-point to 
be 520 C. (125.6° F.) the time of exposure having been four 
minutes, and, although a slightly higher figure has been re- 
corded by other investigators, there is no question but that 
the degree of heat required is low. 56 
WATER-SUPPLY 
" In a moist condition this spirillum retains its vitality 
for months. Koch demonstrated its presence in the foul 
water of a tank in India which was used by the natives for drink- 
ing purposes. It is quickly destroyed by desiccation, as first 
determined by Koch, who found that it did not grow after two 
or three hours when dried in a thin film on a glass cover." 
If the thickness of the film be considerable, or if the drying 
take place on silk threads, the vitality may remain for some 
weeks. (Kitasato.) 
VIABILITY OF THE CHOLERA SPIRILLUM IN WATER 
Babes (1884-85) found the organism alive after seven 
days in Seine water. 
Wolffhugel (1886) found that it may live fifteen to twenty 
days in unsterilized tap-water. He repeatedly found it alive 
after three months, and believes this due to what has been 
termed acclimatization. Rarely the organisms die in the first 
few days. 
Karliniski (1889) found the organism dead after two or 
three days in unsterilized spring-water. 
Hockstetter (1887) found that the organism lives indefi- 
nitely in unsterilized tap-water even when the water contains 
large numbers of other organisms. He found the germs alive 
after an interval of 392 days. 
Nicati and Rietsch (1885) found the spirillum alive in 
sterilized distilled water after twenty days; in sterilized water 
from the port of Marseilles after eighty-one days; in Mar- 
seilles canal-water, thirty-eight days; in sea-water, sixty-four 
days; in bilge-water from an iron steamship en route from 
Japan, thirty-two days. 
According to Giaxa, cholera germs quickly died in water 
containing many other bacteria, and typhoid germs also died, 
but less quickly. Schiller found that cholera germs lived 
fourteen days in a mixture of excrement and urine, and for 
thirteen days in Berlin sewage. Cunningham found them able 
to live four to five days in clear water at room temperature, 
and in dirty water four to nine days. In the latter water, DRINKING-WATER AND DISEASE 
57 
previously boiled, they lived twenty-five days; in garden 
earth, ten to twenty-six days; in the same earth mixed with 
fecal material, six to nine days; in same mixture of earth 
and faecal matter previously cooked-i.e., sterilized-forty- 
seven days. 
Gruber and von Kerner show the power of the cholera 
spirillum to remain alive in river-water, and in that of the 
Vienna city supply, for seven days. 
Christian reports that cholera organisms will live four 
months if frozen with slime.* 
Houston found that they would die rapidly in river-water 
during laboratory storage. At least 99.9 per cent perished 
in one week and none survived three weeks. 
They are more easily destroyed than typhoid bacilli under 
like conditions, f 
These conflicting data are very confusing. The more 
recent ones, such as those reported by Houston, are the most 
reliable. 
According to Boer and Bolton, the cholera spirillum is 
killed by a two-hours' exposure to the following solutions: 
hydrochloric acid, 1 : 1350; sulphuric acid, 1 : 1300; caustic 
soda, 1 : 150; ammonia 1 : 350; mercuric cyanide, 1 : 60,000; 
silver nitrate, 1 : 4000; arsenite of soda, 1 : 400; malachite 
green, 1 : 5000; methyl violet, 1 : 1000; carbolic acid, 1 : 400; 
mercuric chloride, 1 : 10,000; blue vitriol, 1 : 500 
" Experiment has shown the spirillum to be very sensitive 
to the action of acids. Stutzer states that a solution of .05 
per cent of sulphuric acid is fatal to the cholera spirillum in 
fifteen minutes, and a .02 per cent solution kills in twenty- 
four hours. He found that iron pipes could be disinfected by 
sulphuric acid without the metal being sensibly attacked. 
" The most satisfactory evidence that this spirillum is able 
to produce cholera in man is afforded by an accidental infec- 
tion which occurred in Berlin, in the case of a young man who 
was one of the attendants at the Imperial Board of Health 
*Arch. Hyg., lx, 16; also Chem. Abstracts, i., 1036. 
f Metro. Water Board, 5th Research Report, June, 1910. 58 
WATER-SUPPLY 
when cholera cultures were being made for the instruction of 
students." * 
An entirely similar case came under the writer's observa- 
tion, in Paris, while attending the course at the Pasteur Insti- 
tute. One of the students, an Italian, was in the habit of 
constantly smoking cigarettes while at work. He became 
inoculated with Asiatic cholera through laying down his cigar- 
ette in contact with a cholera preparation. He took the 
typical disease and recovered. A friend of the author's reports 
a like instance of infection, observed by him while a student 
in Koch's laboratory. 
Pettenkofer and Emmerich each swallowed pure cultures 
of the comma bacillus, with the result of producing only tem- 
porary diarrhoea, and they thereupon claimed that the germ is 
not to be considered as the cause of cholera. As opposed to 
this, Roux points out that the pure cultures referred to above 
may have been attenuated and very far from the point of 
virulence. Moreover, he shows that, even when truly virulent 
cultures are swallowed, the disease does not surely result. 
Tersely stated, a disease follows when the active patho- 
genic organisms invade the body and find there not only suit- 
able soil for their development but also the absence of antago- 
nistic conditions which prevent their multiplication. 
Dr. David Sommerville,f of Kings College, London, aptly 
says: 
" The intensity of a disease is determined by the number and 
virulence of the particular germ on the one hand and the natural 
or acquired immunity offered by the body to this geim on the 
other. 
11 Many kinds of micro-organisms are permanent residents 
with us, ready to become dangerous at any time that the health 
falls below normal. 
" Highly specialized cells of the human organism possess 
advantages but carry with them serious responsibilities. 
With high specialization has come increased delicacy. 
* Sternberg's " Manual of Bacteriology," 1893. 
f Section viiia. 8th Inter. Cong. Applied Chem. DRINKING-WATER AND DISEASE 
59 
" Pathogenic micro-organisms produce their deadly effects 
through the chemical substances they secrete. These toxins 
call out the production of neutralizing agents. Granted that 
the particular infection is not too intense and the body nor- 
mally healthy for the most part, recovery ensues." 
The President of the National Health Society of England 
some years ago said in an address: "We may lay aside all 
pedantry and mystery talk of epidemic constitution, pan- 
demic waves, telluric influences, cholera blasts, cholera clouds, 
blue mists, and the like terms of art with which an amiable 
class of meteorologists has delighted to cloak their ignorance. 
Cholera is a filth disease carried by filthy people to ■ filthy 
places. It only develops where it finds dirty places, and the 
dirty habit of drinking polluted water and living on a polluted 
soil. Cholera does not travel by air-waves or blasts. We 
drink cholera and we eat cholera, but we cannot catch cholera 
as we catch measles, scarlatina, or whooping-cough. 
" Cholera is always carried. It never travels on its own 
account or by its own conveyance, and it is not half so bad 
a disease as it has been painted by a frightened public. 
" It is stated on the authority of the head nurse that not 
a single case of cholera originated in the hospital of Hamburg 
during the recent epidemic in that city, though the sick were 
often placed two in the same bed and the dead in long rows. 
Amid the gloom and excitement scores of suspects were 
hurried off to the hospital who. were afterwards found to be 
suffering from some other disease. Not one of these persons 
contracted the disease from the cholera patients with whom 
they were forced to associate. It would seem as if the safest 
place at the time of a great epidemic of cholera would be 
where there is the most sickness. All of these statements 
point to the fact that cholera is not infectious, and that the 
danger has been very greatly overestimated." * 
" Pettenkoffer has given the key to the whole situation by 
saying that filth is like gunpowder, for which cholera is a 
spark. A community had better remove the gunpowder than 
* British Med. J. 60 
WATER-SUPPLY 
try to beat off the spark, for, in spite of their efforts, however 
frantic, this may at any time reach the powder, and if it does, 
is sure to blow them to pieces." (Sedgwick.) 
The incubation period of cholera is commonly five days, 
about one-third of the time required to develop typhoid fever, 
but, although more rapid in its attack, the cholera germ is 
fortunately more easily destroyed; therefore the defensive 
measures taken against typhoid fever will abundantly serve to 
protect against the more serious disease. 
Typhoid fever dwarfs all other water-borne diseases from 
the view-point of the sanitary engineer. The possibility of its 
outbreak is always with us; for it does not have to be imported, 
and consequently cannot be efficiently stopped from entry at 
our seaports as is the case with Asiatic cholera. 
Its name " typhoid " is derived from the fact that it for- 
merly was not differentiated from 11 typhus " fever, a disease 
also known as " jail " or " ship " fever and derived from the 
bite of infected lice. 
Typhoid is slow in manifesting itself, the onset of the 
malady occurring about fourteen days after the specific organ- 
isms are swallowed. This period of two weeks is termed " the 
period of incubation." 
A general condition of malaise gradually develops, accom- 
panied by nose bleed, headache, rose spots on the abdomen, 
diarrhoea and with a rise of temperature which is quite 
characteristic, being higher in the evening and lower each 
morning. Serious cases pass into delirium and coma. 
The intestinal discharges and the urine of patients ill with 
typhoid contain great numbers of the bacilli peculiar to the 
fever and it is the admission of such excretions to their food 
and drink that causes the spread of the disease among human 
beings. 
Typical typhoid has followed the swallowing of a pure 
culture of the B. typhosus by a would-be suicide.* 
Typhoid fever is unquestionably more fatal to the negro 
* Arch. Gen. de Med., 1903: 2: 2197, also Jordan's Gen. Bact., 259. 61 
DRINKING-WATER AND DISEASE 
than to the white race, if we judge by the death-rate alone. 
More black people die from the disease per hundred thousand 
living than is noted among the whites; but the thought presents 
itself, is this not really due to the larger case rate arising from 
the greater carelessness of the blacks as a class, and the conse- 
quent opportunity for contact transmission? No answer for 
this question is at hand, but the fact remains that, as Johnson 
has shown, the average typhoid deaths per 100,000 of each 
race living in twenty-four large American cities for 1910 to 1913 
inclusive, stand: White 18.7 and colored 32.1. 
Typhoid is by no means alone in its rating as essentially 
a " human " disease. 
Scarlatina, syphilis and leprosy are among those belonging 
to that class; while, on the other hand, cattle plague, strangles, 
contagious pleuro-pneumonia and fowl cholera are instances 
of animal diseases innocuous to man.* Of course there are 
numerous diseases which attack both man and other animals 
equally well; for instance, the Bubonic plague, which, though 
a rat disease, is communicated to man by the bite of the rat- 
flea. 
The typhoid bacillus is usually one to three microns long 
and from 0.5 to 0.8 micron broad. Its ends are rounded. 
Growth readily takes place at ordinary temperatures in culture 
media, and the colonies do not liquefy the gelatine. Spores 
are not produced. If added to milk it will develop abundantly, 
a property which has been productive of many serious out- 
breaks of the disease. 
Blythe considers its probable normal existence that of a 
saprophyte-i.e., an organism subsisting on decaying organic 
material, f 
There are others who believe, and it is a conceivable belief, 
that the progenitor of the typhoid bacillus is often a sapro- 
phyte, which takes on its pathogenic properties by cultivation 
through successive generations, under favorable conditions 
* " Immunity in Infectious Diseases." Metchnikoff, page 41. 
t Blythe, " Manual of Public Health." 62 
WATER-SUPPLY 
bacillus typhosus, from twenty-four-hour culture on agar. 
(From Hiss and Zinsser's " Bacteriology." Courtesy of D. Appleton & Co.) 
bacillus typhosus, showing flagella. (After Frankel and Pfeiffer.) 
(From Hiss and Zinsser's " Bacteriology." Courtesy of D. Appleton & Co.) 63 
DRINKING-WATER AND DISEASE 
as to light and temperature, and amid suitable surroundings. 
Many illustrations are available, in the world of larger vege- 
tables, of great changes in structure and properties due to 
cultivation under an altered environment. Hansen, the emi- 
nent Danish investigator, dealing with unicellular yeasts, 
has shown that by subjecting these bodies to different environ- 
ments for a number of generations he can from one original 
create many so-called species.* Isolated cases of typhoid 
may be thus accounted for where it would be difficult to sup- 
pose contagion from a previously existing case. 
In a paper before the British Medical Association, H. R. 
Kenwood suggests the possibility of the typhoid bacillus being 
an evolution from the bacillus coli communis, an organism ever 
present in the intestines, and adds that greater changes may 
be artificially induced, both functional and morphological, in 
bacteria than are represented by the slight difference between 
the bacilli in question. 
Klotz observes that: " Of the external influences which 
can be brought to bear on bacteria, alteration of the quantity 
or quality of the food supply plays the most important role, 
and leads to modification of their biological nature. If the 
colon bacillus be grown over an extended time in river water, 
its power to produce indol is diminished or entirely lost." 
" Jenner found that he could revert B. coli capsulatis to 
an unencapsulated form by cultural methods. A remarkable 
difference was noted in the pathogenesis of these two varieties, 
for, as we know, B. coli capsulatis is very pathogenic for white 
mice, but becomes less fatal or even non-pathogenic on losing 
its capsule." f 
Roux and Rodet hold very strongly that the bacillus coli 
communis, when grown under suitable conditions, may become 
pathogenic and produce typhoid. £ Germano and Maurea 
arrive at similar conclusions after careful investigations^ 
* See Bui. N. C. Board of Health, xvi. 100. 
f Am. Pub. Health Asso., 1905, II, 35. 
J Analyst, xxi. 118. 
• § Central, fiir Bakteriologie, xv. 60. 64 
WATER-SUPPLY 
A most disastrous outburst of typhoid occurred during 
June, 1902, at a famous school in Pennsylvania. No previous 
case of the fever had existed at the school and the water supply 
was secured from a deep rock drilled well. The sanitation of 
the entire establishment had been planned by an expert in 
such matters. Upon investigation it was found that a break 
in the sewer pipe had allowed sewage to seep down the out- 
side of the well-casing to its bottom. The lack of a previous 
typhoid history is the interesting feature of this case, for the 
disease was apparently derived from " normal " sewage. We 
must nevertheless remember that the possibility of the presence 
of a typhoid " carrier " (see page 100) is to be reckoned with 
in all such cases. 
Remlinger and Schneider published an article in the Journal 
of the Pasteur Institute, February, 1897, on " The Ubiquity 
of the Typhoid Bacillus," in which they conclude that " the 
bacillus typhosus is distributed in nature outside the human 
body," and they " incline to think that bacilli not pathogenic, and 
indifferent to the serum test, which are encountered in water and 
soil are only definitive types of the bacillus typhosus; at least the 
parentage is evident even if the identity is not absolute." * 
It must not be assumed that those who favor the view of a 
saprophytic ancestry for the typhoid germ have the argument 
all their own way. Far from it. Doubtless the bulk of bac- 
teriologists are ranged upon the other side; f but, while the 
specialists are searching for further light upon this question, 
the evolution, or saprophyte, theory is a good working formula 
for the sanitarian, and upon it he should for the present rest, 
remembering that typhoid fever and filth are very closely 
related. 
"We have yet to learn that the excrement of healthy, 
much less diseased, mammals and birds is altogether harm- 
less to man." t 
* In this connection see an article by F. C. Curtis on " Life-history of the 
Typhoid-fever Germ outside the Body," Albany Med. Annals, xviii. 167. 
t E.g., Sedgwick in J. N. E. Water-Works Asso.,xi. 255. 
t Rideal, " Sewage," page 67. ' DRINKING-WATER AND DISEASE 
65 
The great influence of light upon the growth of the typhoid 
bacillus has been demonstrated by Janowski, who found that 
freshly inoculated gelatine, if kept in the dark, developed 
colonies in three days; if placed in diffused daylight, growth 
occurred in five days; but if the exposure were to direct sun- 
light for six hours, the gelatine became sterile. 
This inability to survive long exposure to sunlight in pres- 
ence of air is not peculiar to the typhoid bacillus. Fortunately 
R.P.I. 
ILLUSTRATING STERILIZING ACTION OF SUNLIGHT. 
for us, such sterilizing action is of wide application, and is one 
of nature's chief lines of defence against overwhelming bacterial 
invasion. A simple illustration, showing the inhibiting action 
of sunlight toward such common bacteria as liquefy culture- 
jelly, may be readily made as follows: Pour some melted jelly, 
previously inoculated with a drop of broken-down culture 
medium, into a Petri dish, upon the bottom of which have 
been pasted letters cut from black paper. When the jelly has 
set, expose the inverted dish for several hours in a cold place 
to the direct sunlight. After exposure, place the preparation 66 
WATER-SUPPLY 
in-the dark, at the usual culture temperature (200 C.). Lique- 
faction will be found to take place only in the portions 
shaded by the paper, and the letters will be found sharply 
countersunk in the jelly. 
Burnett found that the water furnished to Colombo, in the 
island of Ceylon, although not of high quality from a c.hemical 
point of view, rarely contained more than two microbes per 
cubic centimetre. As the supply is from extensive, shallow 
surface-waters, the explanation is offered that nearly complete 
sterilization results from prolonged exposure to the direct 
rays of the tropical sun.* 
All bacteria are not killed by direct sunlight, as, for in- 
stance, B. photometricum.] 
It is generally observed that the number of bacteria in 
river-water is less in summer than in winter, but it must not 
be hastily concluded that this is due to the sterilizing action of 
light. As is shown upon another page, the summer feeders 
are commonly springs, while in winter much impure surface 
washing reaches the streams. Moreover, the action of light 
does not penetrate the water to any considerable depth. 
The thermal death-point of the typhoid bacillus was found 
by Sternberg and Janowski to be 56° C. (132.8° F.), the 
time of exposure having been ten minutes. 
That typhoid bacilli are not completely destroyed by 
exposure to low temperature, the epidemic at Plymouth, 
Pa., in 1885, clearly shows. As we have seen (page 32), the 
outbreak was traced to the dejecta of a single patient, which 
had been thrown upon the frozen ground and snow, during 
the early part of January, and which remained there until 
washed into the stream by the thaw occurring on March 26th. 
During this period the temperature had fallen to 220 F. below 
zero. 
Very similar to the Plymouth outbreak is the one which 
occurred at Windsor, Vt., a town of some 2000 inhabitants, 
* Chem. News, Ixx. 285 
f Wesbrook, J. Pathology and Bacteriology, London, Vol. iii., No. 4. DRINKING-WATER AND DISEASE 
67 
during the spring of 1894. For several weeks the typhoid 
germs remained in a frozen state, without having the potency of 
all of them destroyed. 
As to the longevity of the typhoid bacillus in water, it is now 
generally admitted that such a medium is not favorable to 
its growth, and that although it may survive for varying 
periods, the length thereof depending upon local conditions, 
it rarely, if ever, increases in numbers, and it ultimately 
dies. 
Frankland found that the great majority of typhoid bacilli 
died in raw river-water within a week's time, although a few 
remained alive over thirty days. He laid emphasis upon the 
fact that a factor determining 11 the longevity of pathogenic 
bacteria placed in water, or other unfavorable surroundings, 
is the absolute number in which they are present. 
"Thus among 1000 bacteria from a given source there may 
be some individuals who will resist a particular adverse in- 
fluence, while among 10 bacteria taken from the same source 
there may be none capable of resisting the adverse influence in 
question." * 
Comparing " cultivated " typhoid bacilli, obtained from 
laboratory culture on agar, with 11 uncultivated " bacilli, 
obtained from centrifugal!sed urine secured from the same 
" carrier " that furnished the culture on agar, Houston es- 
tablished the remarkable fact that " uncultivated typhoid 
bacilli die much more speedily in raw-river water than their 
cultivated brethren." f 
In eighteen experiments performed by him, in which he used 
inoculated river-water containing an average of 1,606,673 
" cultivated " typhoid bacilli per cubic centimetre, he found 
that after laboratory storage in glass bottles for one week the 
average number present was only 531 per cubic centimetre, 
a reduction of 99.9 per cent. A few persisted beyond that 
time, but none lived nine weeks. 
* J. Soc. Chem. Ind., 30, 327. 
f 6th Research Report, Metro. Water Board, Nov., 1910. 68 
WATER-SUPPLY 
In the same water inoculated with " uncultivated " bacilli 
direct from a "carrier" the count, at the end of one. week, 
was reduced from 770,000 to 4 per cubic centimetre and at the 
end of two weeks all were dead. 
This demonstration is very far-reaching, showing as it does 
that laboratory data touching the longevity of B. typhosus err 
on the side of safety; it being the " uncultivated " which we 
have to guard against in polluted water. 
Whipple and Mayer have shown that B. typhosus lives much 
longer in water containing dissolved oxygen than in that from 
which such gas is absent; thus showing the lack of benefit 
to be derived from falls and rapids.* 
Hoffmann found the germ to live longer in aquarium mud 
than in the water over the same.f 
Whipple and Lochridge note that a forty minute ex- 
posure to .005 normal HC1 or H2SO4 is fatal to B. typhosus, 
while B. coll requires four times that strength for like 
results. Hence mine water is unfavorable to the life of B. ty- 
phosus. 
Elsdon and Evers find that in artificially carbonated waters 
there is a decided reduction of bacteria. Comparing the 
same water carbonated and uncarbonated there was a reduc- 
tion in the former of 94 per cent in three days and 97.5 per 
cent in seven days, the counts being on gelatin kept three days 
at 210 C.J 
On the other hand, while Young and Sherwood consider 
carbon dioxide as somewhat toxic to B. coli, B. typhosus and other 
bacteria, they believe it is not so much so as to prevent disease 
transmission by means of carbonated beverages. Thus they 
have shown § the following counts per cubic centimetre of B. 
typhosus and B. coli in the same liquid (" pop ") carbonated 
and uncarbonated for the times specified. The carbonated 
liquids were filled at io° C. and 18-pounds pressure. 
* Am. Pub. Health Asso., 1905, ii., 76. 
f U. S. Hyg. Lab. Bui., 35, 177. 
J Analyst, 37, 395. 
§ J. Indus, and Engr. Chem. III., 495. DRINKING-WATER AND DISEASE 
69 
Carbonated 
Uncarbonated 
B. typhosus. 
B. coli. 
B. typhosus. 
B. coli. 
o hours 
200,000 
950,000 
200,000 
950,000 
4 hours 
25,000 
250,000 
200,000 
950,000 
28 hours 
9,000 
20,000 
50,000 
lost 
80 hours 
1,200 
1,300 
6,000 
100,000 
244 hours 
IIO 
900 
900 
40,000 
Pfuhl detected B. typhosus in artificially infected seltzer 
water after twenty-seven days.* 
Mills reported the vital period of B. typhosus to be from 
seven to twenty-one days in Merrimack River water, and 
pointed out that this short period of existence presents the 
probable reason why the fever may be readily carried down a 
river from city to city, while a polluted stream may enter one 
end of a lake, whose waters are changed only after months, 
and a water-supply drawn from the opposite end may be 
much more free from specific pollution.! 
One important point to be remembered is the antagonism 
between the disease germs and the common bacteria so largely 
found in surface-water. While such antagonism has been 
broadly noticed, Percy Frankland has placed the demonstra- 
tion in the following form: He introduced typhoid bacilli into 
deep-well water which was almost wholly free from bacteria, 
into Thames water which contained a large number, and into 
Loch Katrine water, in which the number was intermediate be- 
tween the two. He found that the typhoid bacilli died off more 
rapidly in Thames than in Loch Katrine water, while they per- 
sisted longest in the sparsely populated deep-well water. Thus 
the longevity of these pathogenic bacteria was inversely pro- 
portional to the bacterial population in the waters into which 
they were introduced.^ Hence we see that a relatively pure 
stream, if a rapid one, might carry infection over long dis- 
tances. 
* U. S. Hyg. Lab. Bui., 35, 181. 
t Am. Soc. C. E., xxx., 364. 
J Water and Gas Review, Jan., 1897. 70 
WATER-SUPPLY 
It should be noted here that the foes of the germs of disease 
are by no means confined to the other forms of bacterial life 
that may chance to be their neighbors. Organisms of larger 
growth, grouped under the general term 11 Plankton," serve 
as very energetic destroyers of the pathogenes and greatly 
aid in the sanitary improvement of water. 
Inasmuch as experimental determinations such as the above, 
of the longevity of the typhoid bacillus have been commonly 
made in glass flasks under laboratory conditions, it is interesting 
to note that, for some reason, the data so secured are not 
strictly comparable with the facts as observed in nature. This 
is especially true if the glass vessels used be small ones. In 
commenting upon the divergent results secured by different 
observers, Jordan says: " Substantial accord has resulted on 
two points: first, that typhoid bacilli die out more rapidly in 
unsterilized water than in the same water sterilized by heat; 
second, that when typhoid bacilli are introduced into unsterilized 
water containing little organic matter, their longevity is more 
prolonged than in water charged with considerable organic 
matter." * 
In order to secure conditions more nearly approaching 
those found in nature, Jordan, Russell and Zeit f abandoned 
the usual laboratory methods during their experimental work 
in connection with the Chicago Drainage Canal case and 
employed sacks of either celloidin or of vegetable parchment, 
which, after seeding with B. typhosus, were hung in the 
watercourse under investigation. The problem before them 
was, how long would the typhoid germs live in the water 
of the drainage canal and in that of the Illinois River? 
The sacks, having been filled with the canal or river water 
and seeded with typhoid bacilli, were then so hung as to 
prevent entrance of material floating upon the surface of the 
water. Under such circumstances dialysis would permit 
soluble products of bacterial origin to pass through the parch- 
ment in either direction, while organisms of any kind would 
* J. Infectious Diseases, i., 641. 
t Ibid. DRINKING-WATER AND DISEASE 
71 
be constrained to remain upon whichever side of the dividing 
membrane they originally found themselves. 
Such conditions would serve to protect the imprisoned 
typhoid bacilli from an undue concentration of their own 
products of growth while exposing them to the inhibiting action, 
if any, of the toxins produced by the other organisms present 
in the outside water. 
After extended experimentation along this line these in- 
vestigators concluded that: " Vnder conditions that probably 
closely simulate those in nature, the vast majority of typhoid 
bacilli introduced into the several waters studied perished 
within three to four days. It is theoretically possible that 
specially resistant cells may occur which are able to with- 
stand for a longer period the hostile influences evidently pres- 
ent in the water. Our experiments, however, show that if 
such resistant individuals exist they must be very few in num- 
ber and constitute only a small fraction of the bacilli origi- 
nally entering the water. 
"It is not the intention of the writers to claim that the be- 
havior of typhoid bacilli under the conditions herein described 
is representative of all conditions obtaining in all natural bodies 
of water." 
The reliability of the foregoing experiments has been ques- 
tioned by some observers,* but the subsequent and more 
extended investigations of Russell and Fuller f confirm their 
accuracy. Among the latter's conclusions we note: The 
longevity of B. typhosus in flowing water from Mendota Lake 
ranged from eight to ten days. When exposed to the action 
of sewage bacteria it was reduced to from three to five days. 
When exposed to the diffusible products of sewage bacteria, 
but not to the direct action of the organisms themselves, the 
longevity of the pathogenic germs was increased. 
" The question of longevity is more dependent upon the 
actual contact with sewage types than it is upon contact with 
by-products capable of diffusion through permeable membranes." 
* Engineering Record, 52, 344. 
f Am. Pub. Health Asso., 1905, ii., 40. 72 
WATER-SUPPLY 
" The longevity of the typhoid bacillus in waters is materially 
affected by the germ content of its surroundings. In waters 
highly polluted with saprophytic bacteria, such as is the case 
in sewage, this disease organism is unable to survive for 
more than a few days (three to five in the experiments de- 
scribed), a period of time materially shorter than that which 
is noted in normally unpolluted waters." 
In 1893 Sedgwick and Jackson arrived experimentally at 
the conclusion that typhoid germs would survive many days 
in sewage from which all other germs had been removed by 
filtration, thus showing that the saprophytic toxins were not 
fatal to their existence.* 
Horricks reached a similar conclusion, but he found that the 
germ could not live fourteen days in raw, unfiltered sewage, f 
Pickard claims that less than one per cent of the typhoid 
bacilli can maintain life in a septic tank for the above period.^ 
Frith and Horricks found the bacillus alive in soil after 
seventy-four days; in dry sand after twenty-four days, and in 
peaty soil for a shorter period. § 
Harris states that it can live long in dust and can pro- 
duce true typhoid-pneumonia when inhaled. || 
Dust infection is not unknown. Thus in Nottingham, a 
" dry-closet " town, there was a typhoid outbreak following 
the accumulation of large deposits of night-soil at the city 
refuse depots. Investigation left no doubt as to the source of 
the infection.^ 
In sundry soils, kept moist, B. typhosus will live at least two 
months, without tending to grow or multiply. Such soil 
when converted into dust still contains live germs. The obser- 
vations upon which this is based were not quantitative.** 
From the results of recent investigators " it would appear 
* Chicago Drainage Canal Case , viii., 7971. 
f Rideal, " Sewage," page 66. 
t Ibid., page 63. 
§ Thresh, " Exam, of Water," 139. 
11 Sanitarian, 53, 311. 
J. Roy. San. Inst., xxvii., 533. 
** Brit. Med. J., 1902, 936. DRINKING-WATER AND DISEASE 
73 
that all forms of moist human excrement, dirt, soil, sand or 
gravel favor the viability of the typhoid bacillus. Rullmann 
found that it lived for at least a year and a half in otherwise 
sterile earth and gravel, but in that time had died out in sand. 
The numbers were greatly reduced, however, greater in the 
earth than in the gravel.* 
Wurtz found B. typhosus on the tips of growing vegetables, 
cress, lettuce and radishes, three weeks after sprinkling the 
soil with a watery suspension of the germ.f 
Rullmann J reports the recovery of B. typhosus from 
sterilized soil sixteen months after infection. From un- 
sterilized soil it was recovered after one hundred days. The 
soil, however, did not contain many organisms' of other types. 
Tavel § points out, what has been noted elsewhere, that 
although B. typhosus does not live long in the water of streams 
and lakes yet it can exist for considerable periods in the mud 
upon their bottoms and sides, and he asks attention to the 
consequent danger possibly lurking in the " blind ends " of 
water pipes. 
" The difference in resistance of the typhoid bacillus in 
soils and in liquids under apparently similar physical condi- 
tions may possibly be due to a greater amount of oxygen in 
the soils than in the liquids. This conception is of importance 
as it involves the question of the effect of the almost total lack 
of oxygen in the effluent of septic tanks upon the typhoid 
bacilli which are present in most raw sewages." || 
Dr. Martin, of England, finds that " if the bacillus be 
added to a cultivated soil the micro-organism cannot be obtained 
from such soil after twelve days or so." The shortness of 
this period is not in accord with the observations of sundry 
other writers, but it goes to support the general proposition 
that shallow rather than deep burial is the better way of dis- 
* Pease, Albany Med. Annals. 
t U. S. Hyg. Bui, 35, 184. 
t Centralblal t fur Bakteriologie, xxx, No. 8. 
§ Ibid., xxxiii, No. 3. 
|| Pease, Albany Med. Anal. 
J. Fk. Inst, Oct, 1895, 302. 74 
WATER-SUPPLY 
posing of excreta, for the reason that the upper level of the 
soil is the portion most densely inhabited by the organisms 
likely to inhibit the development of the typhoid germ. As 
compared with the millions of bacteria in the surface layer, 
the soil at the depth of a few feet may be considered as almost 
sterile. 
At a mill at Canton, Mass., in June, 1888, out of 120 men 
some 50 were taken with typhoid. The families of these men 
were not affected. The drinking-water of the mill was from 
a well on the opposite side of a ledge and 54 feet distant from 
a privy-vault, which latter had received typhoid dejecta eight 
months previously. By experimenting with salt, direct con- 
nection by infiltration from vault was shown. Note the time- 
element in this case.* 
Horrocks notes that " soil bacteria (the saprophytes of 
mould and earth) have a destructive effect upon B. typhosus."] 
In his testimony in the Chicago Drainage Canal Case, 
Jordan compared B. typhosus and B. coli as follows with respect 
to their relative abilities to withstand certain adverse conditions: 
" About 50 per cent of the B. coli and 75 per cent of the 
B. typhosus were destroyed by fifteen minutes freezing, after one 
hour 95 per cent of the B. coli and 98 per cent of the B. typhosus 
were killed, and at the end of twenty-four hours over 99 per 
cent of all the organisms had disappeared. 
" Both species resisted temperature up to 450 C. for five 
minutes. And between 450 and 550 C. all but a few indi- 
viduals of each were destroyed. These few individuals, how- 
ever, resist temperature up to 85° C. at which temperature 
all organisms of both species are destroyed." 
Comparison of the resisting powers of the two bacilli 
is important, as it enables us to safely assume the death of 
B. typhosus if B. coli should fail to survive. We may go a 
step further and say that, inasmuch as the spirillum of Asiatic 
cholera is more easily killed than the bacillus of typhoid fever, 
the B. coli will serve as a safety indicator for both. 
* J. New Eng. Water-works Asso., v., 150. 
t J. Roy. San. Inst., 32, 108. DRINKING-WATER AND DISEASE 
75 
As to the diagnostic value to be attached to the presence of 
B. coli in water it may be well to cite a report of a commit- 
tee of the American Public Health Association presented at 
the Washington meeting.* 
"To present a statement of the opinion of the committee 
individually and as a whole, as probably representing the gen- 
erally accepted views of sanitary bacteriologists as to the 
significance of B. coli in water, the following questions were sent 
to each member. The answers were practically as given below: 
" 1. In view of the fact that numerous investigators have 
found B. coli in nature where it could not be directly traced 
to sewage or fecal pollution, do you believe that the colon 
test of water is as safe an index of pollution as it was formerly 
regarded to be? 
11 Yes. 
" 2. Are you of the opinion that the number of colon bacilli 
rather than their presence should be used as a criterion of 
recent sewage pollution? 
11 Yes. 
" 3. To pronounce a water sewage polluted, would you re- 
quire as evidence that B. coli were present in a majority of one 
cubic centimetre samples? 
" Yes, in general." 
In this connection Houston says: "The death of B. coli 
is as sure and safe an indication of the death of the microbes 
of water-borne disease as, to take a homely illustration, the 
death of coarse fish in a water would lead the fisherman to 
assume the coincident destruction of the less hardy forms of 
fish life, e.g., trout." f 
While noting the foregoing it must not be forgotten that the 
presence of a few B. coli may be due to very ordinary causes; 
thus some of them found in a reservoir may have come from 
the excrement of birds or through infection of the water by 
dust from a nearby highway. 
Mr. N. S. Hill advises the writer that sparrows roosting on 
* J. Am. Med. Asso., 1903, page 1292. 
. f Fourth Report Metro. Water Board, June, 1910. 76 
WATER-SUPPLY 
the open well outlets caused the appearance of B. coli in a deep- 
well supply of which he had charge. 
The rating of typhoid fever as a " country disease " has been 
frequently proposed. 
In an address by Dr. J. S. Fulton of the State Board of 
Health of Maryland in 1905,* the speaker quoted data showing 
that the several states could be classified on a decending scale 
"of population density, showing typhoid mortality increasing 
as rural conditions more and more prevail." Thus the per- 
centages of rural populations and the typhoid death-rates 
then were: 
95 per cent population rural 67 per 100,000 died 
87 " " 11 62 
75 " " " 46 " " 
67 11 " 11 --.38 " " 
( C C < ( C ( c c c 
49 42 
Cl Cl Cl C C ( ( 
3° 25 
For Vermont, which is essentially a rural state, only 13 per 
cent of its population being collected in cities, Stone and 
Moat f report that the rate for the urban population is 27.8 
per 100,000 while the rural rate reaches 31.7. 
The census gives for the " Registration States " the follow- 
ing average typhoid death-rates per 100,000 inhabitants for the 
years 1910-1913 inclusive: 
Urban 18.2 
Rural 20.5 
For the state of New York the typhoid rate for 1912 was: 
For the largest six New York cities 10.1 
For the " rest of district " 12.0 
The expression " rest of district " is used to cover that por- 
tion of the sanitary district yet remaining after the cities, 
*N. Y. Bd. Health, 1905. 
f Ground Waters of Vermont, Am. Pub. Health Asso. DRINKING-WATER AND DISEASE 
77 
towns, and villages have been accounted for. It embraces 
the most rural portion of the community. 
More recently this difference has grown smaller and in 1915 
the former relation was reversed for the State of New York, 
the figures being: 
Typhoid rate for the cities of the state 7.95 per 100,000 
Typhoid rate for the entire state 7.7 
The difference observed in many parts of the country between 
the city and country rates must be largely due to the greater 
care exercised in the selection of a water-supply for a city 
as compared with that so frequently displayed in the sinking 
of a country well. It would seem that a due saving of the steps 
of the housewife is what many a well-digger thinks about when 
selecting a site for his well, and he digs it in the most con- 
venient position, without regard to local surroundings. 
The maintenance of the water-supply in a pure state is not 
of itself enough to eliminate typhoid fever, as the disease has 
a number of other channels through which it can be transmitted. 
Infected milk is a common source of typhoid, the infection 
being caused by washing milk cans in infected- water. Given 
a polluted well upon a dairy farm there may be greater danger 
in using its water for washing milk containers than in using 
the water itself for drinking, because milk is a medium in which 
typhoid germs will develop and multiply while water is not. 
The 11 spot-map " of a 11 milk epidemic " is characterized 
by the spots being likely to appear along a particular milk- 
route; moreover, young children figure largely in the list of 
cases, and the outbreak is likely to be of an " explosive " type. 
Typhoid distribution by flies has often been the real origin 
of epidemics of the fever when the water supply has been blamed 
as the cause. 
At 700 F., or over, the life cycle of the ordinary house fly 
can be completed in ten days. Each female may lay four or 
five batches of eggs containing somewhat more than 100 eggs 
in each. Howard estimates that the immediate descendants of 78 
WATER-SUPPLY 
a single pair of flies might number twelve millions and weigh 
800 pounds. Flies take only liquid food. When feeding on 
dry material they regurgitate liquid material from their 
" crop " whereby solution of the food is secured and then the 
" enriched mixture " is sucked in. The fly's feet " have each 
a two-lobed pad, covered with a dense growth of very minute 
tubular hairs, from which a sticky secretion can be exuded; 
it is by means of this apparatus that flies are enabled to walk 
on smooth surfaces in any position." 
(After Howard.) 
Dr. Graham-Smith * sets forth some of the dangers attend- 
ing the habits of the fly and shows that " infected flies not 
only carry bacteria on their bodies and limbs, thereby con- 
taminating substances over which they walk, but distribute 
bacteria which they have ingested, by means of vomit and fecal 
deposits." He shows that, while non-spore-bearing bacteria 
survive at the most only twenty-four hours on the limbs, flies 
nevertheless infect substances over which they walk with 
such organisms for several days by means of a fluid which 
* " Flies in Relation to Disease," 1913. DRINKING-WATER AND DISEASE 
79 
they regurgitate from their crops. He also shows that " the 
majority of the non-spore-bearing bacilli pass through the 
intestine and are in living conditions in the fecal deposits." 
11 Flies carry in and on their bodies very large numbers of 
bacteria, many of which are fecal types, and these are more 
numerous in flies caught in congested or dirty areas. Patho- 
genic bacteria or allied types have been isolated from wild city 
flies." 
Wherever human excreta are exposed to the reach of flies 
as in unscreened privies, there is always danger of such material 
being transported to the neighboring kitchen, with consequent 
infection of food. This form of typhoid distribution was the 
source of the serious outbreak of the fever among the American 
troops during the Spanish War. 
Dr. L. 0. Howard calls attention to the fact that, " during 
that war, in the temporary camps sufficient attention was not 
paid to protecting the latrines from the visit of flies, and in 
such camps typhoid abounded, while in camps adopting such 
precautions, especially such permanent camps as the Presidio, 
typhoid was almost unknown." * 
In a typhoid epidemic reported by Dr. Hamilton: "The 
streets in which the sanitary arrangements are worst had the 
largest number of cases, irrespective of the poverty of the 
inhabitants. 
" Flies caught in two undrained privies, on the fences of two 
yards, on the walls of two houses, and in the room of a typhoid 
patient were used to inoculate eighteen tubes, and from five 
of these tubes the typhoid bacillus was isolated." f 
In cities and sewered towns where the yard privy is less 
commonly found, fly infection plays a less conspicuous part; 
nevertheless, lack of house plumbing is far from rare, even in 
large places, and " fly typhoid " epidemics are still in evidence. 
The " spot map " of such an epidemic is characterized by 
irregularity in the location of cases. The " spots " are found 
* Pop. Sci. Monthly, Jan., 1901. 
f J. Am. Med. Asso., Feb. 28, 1903. 80 
WATER-SUPPLY 
in the poorer parts of the town, where absence of plumbing and 
the carelessness and ignorance likely to exist in such localities 
allow of the exposure of both excreta and food to fly invasion. 
Typhoid from infected shell-fish follows from the fattening 
or growing of them in polluted waters. Unfortunately some 
of the localities whence wholesome oysters have been taken in 
the past are now exposed to serious sewage inflow with the result 
that the shellfish taken from such beds carry with them a men- 
ace to health. Still more objectionable is the practice of fat- 
tening oysters for the market by transplanting them to the 
mouth of some stream if such stream be a polluted one, as 
sometimes happens. 
It is a fortunate fact that oysters close their shells for the 
winter, hibernate in short, and consequently are relatively 
safe during the cold months. 
The remedy against the danger of shellfish infection is a 
legal one, namely, the closing of polluted waters to oyster 
culture and the proper protection of established oyster-beds 
from sewage inflow. 
The securing of protection against the entrance of disease 
through any of the channels named should, however, be sup- 
plemented by supervision of general hygienic conditions as 
well, otherwise the resisting powers of the human organism 
may be lowered and left unable to successfully oppose the in- 
vading germs, should they come from some unexpected source. 
The experimental results of Charrin, Roger, Pasteur and 
others,* bear directly upon the relation of insanitary surround- 
ings to the development of typhoid, and also upon the fact 
that the effect of physical exhaustion, whether due to cold, 
fatigue, hunger or other cause, is to predispose to the contrac- 
tion of disease. 
The following is quoted from the Bulletin of the Indiana 
State Board of Health: " On one of the farms the well became 
infected, and of twelve men who drank water from it during 
* See Sternberg's Bacteriology; also J. San. Inst., xvi., 487. DRINKING-WATER AND DISEASE 
81 
one day, eight acquired typhoid fever within two weeks. It 
is interesting to note the fact that the eight men who were at- 
tacked were very tired and quite worn out when they drank 
the infected water." 
It is to be presumed that the water-closet system is the one 
which most quickly and efficiently carries away dejecta from 
the premises, and we should expect a corresponding benefit to 
be reflected by the typhoid statistics. 
The following record for the city of London * covering houses 
possessing three forms of sewage disposal speaks for itself: 
Houses with privy system, 1 case of typhoid for every 37 
houses. 
Houses with pail system, i case of typhoid for every 120 
houses. 
Houses with water-closet system, 1 case of typhoid for every 
558 houses. 
Of course some allowance for " secondary " typhoid, due to 
unsanitary habits, must be applied to the above data. 
Considerable discussion, pro and con, has been raised 
with reference to the influence of " sewer-air " upon the spread 
of typhoid fever. 
W. H. Horrocks f summarizes his experimental investi- 
gation as follows: 
" Specific bacteria present in sewage may be ejected into 
the air of ventilation pipes, inspection chambers, drains and 
sewers by (a) the bursting of bubbles at the surface of the 
sewers, (&) the separation of dried particles from the walls of 
pipes, chambers and sewers, and probably by (c) the ejection 
of minute droplets from flowing sewage." 
Professor Winslow strongly opposes this view of Horrocks 
and advances experimental evidence in support of his objec- 
tion; but there is another point from which the question may 
be viewed. Even though we grant the absence of pathogenic 
* " The Purification of Sewage," Barwise, 
f J. Roy. San. Inst., May, 1907. 82 
WATER-SUPPLY 
germs from sewer air, yet surroundings which offend the senses 
indirectly lead towards disease by a lowering of resisting power, 
just as vice versa the placing of a patient amid pleasant sur- 
roundings aids in the accomplishment of a cure. 
There is a direct relation between some forms of filth and 
disease, though such relation may not be always manifested: 
and there is an indirect relation between disease and all forms 
of filth that are appreciable by the senses, and this latter indirect 
relation is the more common of the two. 
Consideration must be given to the feelings and views of 
those whose beliefs are centered upon a close relation between 
any form of sewage pollution and disease, even though the sani- 
tarian may not fully endorse the opinions of such persons from 
a purely scientific standpoint. 
11 Faith cures " are recognized as really curative in some 
disorders, and likewise a strong belief in the bad results likely 
to arise from certain existing surroundings, held to be objec- 
tionable, must be allowed to have depressing effects upon the 
normal resistance to disease invasion. 
With the advance in the inspection and care of water sup- 
plies it must of necessity follow that they will become less and 
less a causative factor of disease, and the typhoid of the future 
will, relatively, be more frequently traced to other sources. 
It is wise to call attention to the prevalence of mild and 
unrecognized cases of typhoid fever, and to the greater danger 
arising from them, and it might be added that such cases are 
dangerous to the patient as well as to the neighbors, by reason 
of the lack of care given them and the consequent exposure 
to relapse. 
As to the likelihood of the disease spreading throughout 
a household by " counter infection," that will largely depend 
upon the class of people among whom the fever primarily 
appears. With filthy people the danger is material, but among 
those of cleanly habits, who are alive to the necessity of fol- 
lowing sanitary instruction, there is small reason to fear an ex- 
tension of the disease. DRINKING-WATER AND DISEASE 
83 
" Secondary " and " residual typhoid " are terms intended 
to cover those cases arising from causes other than the water 
supply; the first due to the filthy habits of people who may 
handle food, and the second to other causes not specifically 
classified. At the present moment it seems that, for American 
communities, an average yearly mortality of about fifteen 
per hundred thousand inhabitants must be counted as due to 
these two causes. 
The number is large and it is hoped that it may be lessened 
by the extension of a better popular knowledge of the origin 
of typhoid and of the methods of guarding against its occurrence. 
Typhoid in general, and therefore residual typhoid as well, 
has been lessened for the State of New York, as is shown by 
the following record (compiled by Theodore Horton): 
ANNUAL TYPHOID DEATH RATES PER 100,000 POPULATION (STATE 
OF NEW YORK) 
1895 25.4 
1896 22.6 
1897 19.4 
1898 25.6 
1899 22.4 
1900 26.7 
1901 23.4 
1902 17-4 
1903 21.5 
1904 20.9 
i9°5 iQ-2 
1906 19.0 
1907 19.8 
1908 16.0 
i9°9 15 i 
1910 15.0 
1911 14.o 
1912 11.8 
i9I3 IO-5 
1914 8.8 
1915 7-7 
A list of typhoid rates for sundry large cities of the world 
will be found on pages 39 and 40. 
Dr. Rideal of London has suggested that fittings be installed 
in each house, such as Pasteur filters, heat sterilizers and the 
like, which same, being placed under municipal control, sterility 
of the public water supply used for drinking could be insured,* 
with consequent reduction of the typhoid rate. 
Such a procedure would not be practicable, for the reason 
that the ignorant portion of the community must be furnished 
with a water that is " safe from the faucet." 
After all sources of possible danger have been examined, 
it must be admitted that outlying isolated cases of typhoid 
* J. Roy. San. Inst., xxvi., 679. 84 
WATER-SUPPLY 
fever arc often difficult to explain; but it should not be for- 
gotten that the disease does not manifest itself until a con- 
siderable time after infection, the incubation period being 
usually about fourteen days, and therefore the possibility of 
its having been imported should be always borne in mind. 
" After the reception of the infection, there is, in all com- 
municable diseases, an interval during which the patient 
remains in apparent health, or perceives at the most some 
languor. This period lasts from one to five days in the case 
of cholera. For typhoid fever its duration varies from nine 
days to three weeks. The latter disease begins so gradually 
that the patient generally does not come under the observa- 
tion of a physician until he has had the fever for several days. 
If water infected by typhoid-fever dejecta were to be drunk 
by a considerable number of persons August first, the first 
case would appear about the ninth or tenth, and fresh cases 
would continue to appear until the twenty-third. There would 
be more on the fourteenth or fifteenth than at any other time. 
The deaths would nearly all occur the next month-Septem- 
ber. These laws of development are of great aid in discover- 
ing the cause of brief epidemics, by indicating the period in 
which it must have been common to all the persons attacked." 
(E. J. Matson.) 
An argument always advanced against the proposition that 
a typhoid epidemic in a town is to be accounted for by the use 
of a contaminated water-supply is that only a few of the in- 
habitants are attacked, while all use the water. Why should 
the majority escape? For full discussion of the wide subject 
of " immunity," thus introduced, the reader must be referred 
to the extensive monographs written thereon; but let it be here 
said that immunity depends at least in part upon an inherited 
or acquired tolerance to the toxic products of pathogenic bac- 
teria. 
This question of "tolerance" is of wide application and 
affects many forms of life. Thus, according to Metchnikoff: * 
" Whilst the lower organisms are refractory to bacterial 
* " Immunity in Infectious Diseases," pages 21 and 24. DRINKING-WATER AND DISEASE 
85 
toxins which in quite small closes are capable of killing man 
and the higher animals, many micro-organisms manifest a 
special sensitiveness to certain fluids of animal origin. The 
blood of the rat contains an organic base capable of killing 
and dissolving a considerable number of anthrax bacilli." 
Yet it is possible " by successive cultures to accustom the 
anthrax bacillus to an existence in the pure serum of the rat. 
The bacillus protects itself against the toxic action of the 
serum by surrounding itself with a thick sheath composed of 
a kind of mucus which fixes the toxin of the rat's blood and ren- 
ders it harmless. 
" Stentors kept for two days in a weak solution of corro- 
sive sublimate (.00005 per cent) acquire a tolerance to a dose 
of this poison four times as great as the lethal dose for indi- 
viduals previously kept in pure water." 
It is very possible that those people who use a polluted 
river water, and are therefore daily exposed to invasion by greatly 
attenuated typhoid organisms, may gradually acquire an im- 
munity against even virulent bacilli, which protection would 
not be possessed by the visiting stranger. 
According to the theory of phagocytosis advanced by 
Metchnikoff, invading pathogenic organisms are seized upon 
by the guarding leucocytes of the blood, and destroyed by a 
process of assimilation, provided " the captors are not paralyzed 
by some potent poison evolved by their prisoners, or over- 
whelmed by their superior vigor and rapid multiplication." 
" Micro-organisms, after their entrance into a refractory 
animal, are not eliminated by any of the excretory channels 
which serve for the elimination of many of the soluble poisons." * 
Invasion of a susceptible animal by a single disease-germ may 
prove fatal, as has been shown by Cheyne, who experimented 
upon guinea-pigs with anthrax; f but if any considerable 
degree of vital resistance be present, the " bacterial dose " 
* " Immunity in Infectious Disease," page 46. 
t See also " Principles of Bacteriology," by Hueppe, Trans, by Jordan, page 
152. 86 
WATER-SUPPLY 
may have to be very greatly enlarged to produce observable 
effects. Thus the above investigator found that 11 for rabbits 
the fatal does of the microbe of fowl cholera is 300,000 or more, 
that from 10,000 to 300,000 cause a local abscess, and that less 
than 10,000 produce no appreciable effect." He found 225,000,- 
000 of the Proteus vulgaris fatal to rabbits, but that less than 
9,000,000 gave an entirely negative result. 
Another interesting point that arises in this connection is 
the wide difference between the intensity of the attacks in- 
duced by a " virulent " and an " attenuated virus." It is well 
known that, if the conditions attending the cultivation of a 
pathogenic microbe be unfavorable to its ready growth, if they 
be just short of the death-point, if the germ be obliged to 
struggle for existence through successive generations, the 
result is an organism of less vigorous constitution, and one 
capable of producing only a fraction of the amount or intensity 
of " toxin " elaborated by its sturdy progenitor. 
Inoculation with such " attenuated virus " might be fatal 
to the very susceptible, but a larger number of the resistant 
portion of the community would escape, and the great majority 
of all cases occurring would be designated as " mild." 
Typhoid epidemics following the use of polluted river water 
are commonly mild in character, while isolated country cases 
of the disease are often severe. Such has been the writer's 
experience, and a reasonable explanation would seem to be that 
while the adverse conditions of river carriage supply abundant 
opportunity for the pathogenic germs to either die, or to at 
least lose a portion of their virulence, the conditions governing 
the supply of water from an infected well usually admit of a 
shorter period of time to elapse between the entrance of pol- 
luting material and the drinking of the water. In other words, 
the struggle for existence during stream transmission will 
cause a decrease in the " poisoning power " of the typhoid 
germ, which decrease will vary directly with lapse of time, 
and will consequently be a function of both distance and velocity 
of flow. Bacilli, therefore, which started on their journey in 
vigorous condition might be considered as arriving at the point DRINKING-WATER AND DISEASE 
87 
of invasion in a state so enfeebled as to be incapable of pro- 
ducing a " normal " type of the disease. 
Such a proposition was advanced by the writer some years 
ago in a case involving the question of water-borne typhoid, 
and the following data are offered in support of the argument. 
There arose an opportunity to study an outbreak of typhoid 
fever occurring among people who used waters of widely differ- 
ent characters. The public supply of the city was from a stream 
which received the sewage of a town a few miles away, and many 
cases of typhoid occurred among those who exclusively used 
such water. Throughout the city there were numerous 
grossly polluted domestic wells, and fever cases were plentiful 
among families using water from no other source. As was to 
have been expected, a still larger number of cases of the disease 
were found where the water drunk was from both the Wells 
and the river. A careful house-to-ho'use visitation elicited the 
following data. The division of the cases into " light " and 
" severe " rests upon the statements of the attending physicians. 
Water Used. 
Number of 
Cases. 
Light Cases. 
Severe Cases. 
Ratio of Severe 
Cases to Total 
Cases, Per Cent. 
Stream 
61 
44 
17 
27.80 
Well 
46 
20 
26 
56.50 
Mixed 
132 
72 
60 
45-45 
Data such as the above are very difficult to secure, and we 
are usually forced to rest satisfied with mortality returns alone, 
but even from such unsatisfactory material our proposition 
receives no small measure of support. 
In a report prepared for the Parliamentary Bills Committee 
of the British Medical Association, Hart gives sundry figures 
for a number of British epidemics of typhoid fever. It is 
possible to pick out from his collection of facts 108 instances 
where the disease could be traced to contaminated wells upon 
the one hand, or to polluted reservoirs or streams upon the 
other. The table here following gives the average death-rates 
for the two classes of epidemics, and there are added the figures 88 
WATER-SUPPLY 
for milk epidemics also, although such are foreign to the present 
discussion. 
Average Death-rate, 
Per Cent. 
75 epidemics due to well waters 
11.83 
33 epidemics due to stream and reservoir waters 
9 85 
20 epidemics due to milk 
12.79 
In speaking with the writer upon this general subject, a 
Philadelphia physician of large practice said that his typhoid 
death-rate in private practice was about 5 per cent, but that 
upon one occasion it was much higher. He was called to 
attend an outbreak of the fever occurring among a considerable 
number of Russian sailors, who had been sent to Philadelphia 
to man a battleship then building at Cramp's shipyard. These 
men drank water from a well which was afterwards proved to 
have been grossly polluted. The typhoid death-rate resulting 
among them was about 20 per cent. 
Such figures as have been given certainly tend to show that 
greater or less concentration of polluting material and longer 
or shorter exposure of the typhoid germ to unfavorable sur- 
roundings must be held to account in part at least for the vari- 
ability observed in the intensity of the disease. 
It is an odd fact that when typhoid is not very prevalent 
in a city the death-rate is often high-most of the patients 
die; but when the fever is really widespread and there is an 
epidemic, most of those attacked get well. That merely 
shows, of course, that when there is but little typhoid the cases 
are not reported. Physicians will report cases during an 
epidemic, but they will often not report them during a non- 
epidemic period unless they be fatal. 
Thus for the city of Columbus, Ohio: 
Cases Reported. 
Deaths. 
Fatality, % 
January, 1904 
725 
35 
4 83 
February, 1904 
798 
94 
11.78 
March, 1904 
83 
33 
39 76 
April, 1904 
28 
9 
3i 14 89 
DRINKING-WATER AND DISEASE 
Typhoid fever is essentially an autumn disease in this lati- 
tude. Thus we fin/d the following returns of the N. Y. State 
Department of Health for the year 1912: 
The number of deaths from typhoid fever, as reported by 
months, was as follows: 
Urban. 
Rural. 
Urban. 
Rural. 
January 
86 
21 
July 
68 
II 
February 
59 
II 
August 
i°5 
16 
March 
56 
12 
September 
115 
22 
April 
57 
II 
October 
113 
30 
May 
.. 66 
18 
November 
77 
12 
June 
56 
II 
December 
61 
18 
United States census records show that during the year 
of 1910 there were 12,673 typhoid deaths in the " Registration 
Area." These deaths were distributed as follows over the 
twelve months of the year: 
January 728 
February 673 
March 788 
April 652 
May 578 
June 626 
July 832 
August 1,460 
September 1,917 
October 1,858 
November I>53° 
December 1,041 
12,673 
Data such as the foregoing, and others that might be given, 
indicate the autumn to be normally the " typhoid season." 
Temperature conditions have much to do with the seasonal 
distribution of typhoid fever. 
If we remember that summer is the time when flies are most 
numerous, and also bear in mind that each developed typhoid 
case is but an additional centre from which yet another may 
arise, we can see how likely it would be that the summer inocu- 
lations of the fever would swell the autumn death reports. 
The observation has been made that cities which take their 
water from polluted rivers are prone to be exceptions to the 90 
WATER-SUPPLY 
autumnal typhoid rule and exhibit their outbreaks of the 
disease in winter or spring. 
The illustration given herewith shows the amount and 
monthly distribution of typhoid fever at Albany, N. Y., 
during periods before and after the establishment of the city 
filter plant. 
It will be observed that the improvement in the quality 
of the public water has done more than simply decrease the 
Cases before .filtration. 
Cases after, filtration, 
MONTHLY DISTRIBUTION OF TYPHOID FEVER AT ALBANY N. Y. 
(Courtesy of F. R. Lanagan, City Engineer) 
amount of typhoid; it has altered the time of its maximum 
intensity as well. What typhoid yet remains in Albany is 
now an autumn disease and has lost the winter high-point 
which it formerly possessed. 
The chart indicates how the " water-borne " portion of the 
annual typhoid enveloped the " residual " and by its greater 
volume obscured the latter and changed the average date of 
incidence. 
The installation of filters at Providence, R. I., produced DRINKING-WATER AND DISEASE 
91 
an entirely similar result upon the typhoid statistics in that 
city. The improvement of the water resulted in lowering 
the typhoid rate and in changing its date of maximum occur- 
rence from December to October. 
Ruediger* shows that both 11 B. coli and B. typhosus dis- 
appear much more rapidly from polluted river water during 
the summer months than when the river is covered with ice 
and snow." 
Houston in his seventh research report to the London 
Water Board gives in great detail data showing that " tempera- 
ture is an important factor as regards the vitality of the 
typhoid bacillus in river water, and the lower the temperature 
the longer does the bacillus survive." He found that in raw 
Thames water, which contained 103,328 typhoid bacilli per 
cubic centimetre, the average reduction in number of germs 
per cubic centimetre was as follows for a storage of one week 
at the temperatures given: 
320 F 47,766 remaining, or 46.22 per cent 
4i°F 14,894 11 " 14.4 
5o°F 69 " " .06 " 
64-4° F 39 " " -03 " 
80.6° F 19 " " .01 " 
Thus we are to expect the B. typhosus to live its life more 
rapidly and vigorously during warm weather and to quickly 
pass to its death with the assistance of other forms of life then 
existing in plenty; while under the influence of cold it would 
appear to assume a sort of hibernating state and thereby 
greatly prolong its existence. 
The above facts serve to explain why the maxima of typhoid 
outbreaks, caused by polluted river waters, are pushed forward 
into the colder months of the winter and early spring. 
Typhoid fever and height of ground water was a relation 
commented upon by Dr. Michel, of Chaumont, France.f In 
* J. Am. Pub. Health Asso., June, 1911. 
t " Influence de 1'eau potable sur la sante publique." Paris, 1889. 92 
WATER-SUPPLY 
1855 he observed that typhoid, which was epidemic in the above 
place, varied in number of cases and in intensity inversely as 
the quantity of water in the public wells. 
AVERAGE FOR 12 YEARS. 
FULL LINE= TYPHOID. DOTTED LINE = INCHES OF GROUND ABOVE WATER IN WELL. 
COINCIDENCE OF PREVALENCE OF TYPHOID FEVER AND LOWNESS OF WATER IN 
WELLS. (STATE OF MICHIGAN.) 
Latham, in speaking upon this point, says: 
" No great variation in the vertical rise and fall of subsoil DRINKING-WATER AND DISEASE 
93 
water is the healthier condition. Dry summers invariably mark 
unhealthy years. Typhoid fever occurs after the autumn rains. 
" All the great epidemics of typhoid have occurred in years 
when the ground-water was especially low, and after a slight 
rise in the same." 
He especially refers to the year 1741, " which was one of 
the driest periods known, and the most unhealthy period in 
many places that has ever been recorded.'.' * 
A reasonable view of the situation is that, as the water- 
surface lowers in a well, the base of the cone of drainage, whose 
apex is at the water surface, is extended, and consequently 
more widely situated points of pollution are embraced within 
its influence. 
Perhaps the most exhaustive examination of the relation of 
the height of ground-water to the prevalence of typhoid fever 
that has been made in America is to be found in the work of 
the State Board of Health of Michigan. 
Observations have been made by that board during a 
period of many years, and the height of the water in a repre- 
sentative well near the centre of the State has been regularly 
measured. The results, graphically shown herewith (see page 
92), indicate in a marked manner that increase of typhoid 
and lowness of water in wells move in practically the same 
curve of variation. 
Whether the exhaustive study of facts does or does not 
support the view that the relation of typhoid fever and rainfall, 
so far as ground-water is concerned, deals with the question 
of low ground-water rather than with fluctuations in its ver- 
tical height, it admits of ready illustration that marked relation- 
ship certainly exists between this disease and the sudden influx 
of storm-waters, flooding the polluted foreshores of smaller 
rivers. We have seen such a case in the epidemic of typhoid 
in the valley of the Tees, page 26. 
As cities increase in size there are introduced into the total 
death-rate disturbing factors that must be considered in com- 
* J. Roy. San. Inst., 31, 457. 94 
WATER-SUPPLY 
prehensive study. Thus the influence of simple crowding is 
well illustrated by the following statistics for various London 
districts: * 
Mean 
Death-rate, 
1885-91- 
Districts with a density of under 40 persons per acre.... 15.27 
from 40 to 80 19.04 
11 11 from 80 to 120 19.24 
from 120 to 160 22.60 
" " over 160 23.88 
County of London, with a density of over 57 19.90 
In view of the intensely practical spirit of the age, let us 
consider the question, 
Does pure water pay ? 
To abandon an existing water-supply system, or to purity 
the polluted water that it furnishes, always involves the outlay 
of much money, and the city taxpayer has the right to inquire 
whether or not the benefit derived is a fair equivalent for the 
cash expended. 
Typhoid fever is doubtless, to a large extent, a preventable 
disease, but the means of prevention, in the shape of great 
public works, are expensive, and the question is very properly 
asked, Do these works pay? Can we afford to save the 
typhoid victims? 
Before answering let us note that the age at which typhoid 
fever most commonly occurs, as shown by the above chart, 
is the very prime of life. 
According to Rochard, the economic value of an individual 
" is what he has cost his family, the community, or the State 
for his living, development, and education. It is the loan 
which the individual has made from the social capital in order 
to reach the age when he can restore it by his labor." 
The statement of this value, in form of money, is a difficult 
matter, which has been variously settled by sundry investi- 
* Engineering News, Dec. 28, 1893. DRINKING-WATER AND DISEASE 
95 
gators. Chadwick considers an English laborer equivalent to 
a permanent deposit of £200 (say $980). Farr gives £159 
(say $780) as the average value of each human life in Eng- 
land. A French soldier is rated as worth 6000 francs (say 
$1200). 
In view of the fact that typhoid fever selects by far the 
greatest number of its victims from among those in the very 
SHOWING TYPHOID DEATH-RATE AT DIFFERENT AGES, FOR THE STATE OF 
CONNECTICUT. * 
prime of life, to the relative exclusion of the very young and 
the very old, it will be reasonable to follow the figure fixed 
upon by E. F. Smith, and place the loss caused the com- 
munity by a death from typhoid at $2000. This will be 
noticed to be less than half the figure so frequently referred to 
in the courts of this State as the value of a human life. 
For the sake of illustration, let us consider the tax formerly 
* Rep. Conn. Board of Health, 1895, p. 165; see also a similar chart in the 
Report for 1903, xxi. 96 
WATER-SUPPLY 
levied annually by typhoid fever upon the city of Albany, 
N. Y.* 
The population of Albany is about one hundred thousand, 
and from statistics given in the reports of the State Board of 
Health the deaths due to typhoid fever in Albany formerly 
averaged seventy-five for the year. 
Rating the money value of each life at the figure given 
above, this death-rate would mean an annual pecuniary loss 
to the city of $150,000. 
Funeral expenses are variously extimated at from $20 to 
$30. Should we accept the intermediate value of $25, this 
item would cause $1875 to be added to the above sum, thus 
raising the total direct loss through death to $151,875. 
But typhoid fever does not always kill. Its mortality rate 
is commonly quoted at about 10 per cent. For the present 
purpose, should we assume nine recoveries for each death 
from the disease, and place 43 days as the period of convales- 
cence (the average of 500 cases at the Pennsylvania Hospital), 
we should have a term of 29,025 days as representing the time 
lost, per year, by the 675 persons who have the fever and 
recover. Thus an annual loss of over seventy-nine years had to 
be borne by the city's capital of productive labor. 
This great amount of enforced idleness, when translated 
into money value, should very properly be added to the death- 
loss above estimated. 
Fixing the rate of wages at $1 per individual per day-a 
very low figure, considering that the bulk of typhoid patients 
are in the very prime of life-there formerly was a loss of $43 
of wages for each recovery, or a total yearly loss for the city from 
this item of $29,025. The cost of nursing and doctors' bills 
equals at least $25 per case, which is a very low estimate, thus 
adding the further amount of $16,875 to the gross sum. Ex- 
pressed in tabular form, this yearly tax formerly imposed by 
typhoid fever upon the city of Albany is given below, and, 
upon a most conservative estimate, it was practically $200,000, 
* Albany now possesses a modern and efficient filter-plant. See page 144. 
The typhoid rate for 1915 was 12 per 100,000. DRINKING-WATER AND DISEASE 
97 
or $2 a year for each man, woman, and child in the city or a 
yearly tax of $10 for every family of five persons. 
75 deaths at $2000 each $150,000 
75 funerals at $25 each 1,875 
Wages of 675 convalescents during 43 days at $1 per 
day 29,025 
Nursing and doctors' bills for 675 convalescents at 
$25 each case 16,875 
Total tax formerly levied annually by typhoid fever 
upon the city of Albany $197,775 
Dr. V. C. Vaughn estimates the annual number of typhoid 
cases in the United States as half a million, with 50,000 deaths, 
and he places the annual cash loss resulting from the same at 
$90,000,000. 
It can readily be seen that public works which eliminate 
a reasonable fraction of this great tax pay for themselves in the 
course of a few years, even though they were originally expensive. 
" In Pennsylvania the reduction of the typhoid fever death- 
rate has been marked in the ten-year period since the State 
Department of Health was organized and has resulted in the 
saving of thousands of lives to the State. 
DEATHS FROM TYPHOID FEVER AND LIVES SAVED IN 
PENNSYLVANIA, 1906 to 1914, Inclusive. 
Year. 
Typhoid 
Deaths. 
Rate per 
100,000 
Calculated 
Deaths. 1906 
Rate Obtain- 
ing. 
Saving as com- 
pared with 
1906. 
1906 
3917 
54-8 
1907 
3538 
48.6 
3989 
451 
1908 
2450 
32-4 
4064 
1,614 
1909 
1712 
22.7 
4MO 
2,428 
1910 
1892 
24.6 
4216 
2,324 
1911 
1716 
21.9 
4291 
2,575 
1912 
1310 
16.5 
4367 
3,057 
1913 
1470 
18.1 
4443 
2,973 
1914 
1071 
13.0 
45i8 
3,447 
Total saved 
18,869 98 
WATER-SUPPLY 
" In 1905, when the Department began its operations, Penn- 
sylvania was a typhoid ridden state. The purification of the 
water supplies and the improvement in general sanitary condi- 
tions have been factors of tremendous influence in bringing 
about this marvelous reduction in the number of deaths from 
disease so often water-borne in origin." * 
A further item of cost appears to be coming in the near 
future, namely, that of " damages for the serving of unsafe 
water." 
In April, 1915, the Supreme Court of New Jersey decided 
in favor of a plaintiff, awarding him expenses and loss of time 
because of the illness of three of his children from typhoid 
contracted from water supplied by the local water company. 
To quote from the decision: 
" It must be borne in mind that the defendant company 
was in the water-supply business for profit. The plaintiff had 
paid for the supply which he was to receive in advance. 
Hence, it became the duty of the defendant company to give 
to the plaintiff water fit for domestic purposes, including 
fitness for drinking. 
" Water is a necessity of life and one who undertakes to 
trade in it and supply customers stands in no different posi- 
tion to those with whom he deals than does a dealer in food- 
stuffs. He is bound to use reasonable care that whatever is 
supplied for food or drink shall be ordinarily and reasonably 
pure and wholesome." 
Citation was made of a decision f " that pure and whole- 
some water necessarily means such as is reasonably free from 
bacteria and coli, or any other infection or contamination which 
renders water unfit for domestic use and unsafe and dangerous 
to individuals." 
The court held that making it " reasonably appear that the 
drinking water was the probable efficient cause of the typhoid 
fever," was all that the plaintiff could be required to do. 
In Wisconsin the supreme court has handed down a decision 
* Pennsylvania Health Bulletin, Aug. 1915. 
f Peffer v. Penn. Water Co., 221 Penn. 578; 70 Atl. Rep., 870. DRINKING-WATER AND DISEASE 
99 
in favor of an employee who died from drinking polluted water 
furnished him by a lumber mill company.* 
In a paper by J. J. Hoppes he very tersely referred to 
the fact that the comparatively few deaths from typhoid 
fever in the army during the Spanish war had stirred up the 
newspapers to much comment and criticism, while they scarcely 
mention that 40,000 deaths from this disease occur annually 
in the United States, and many times that number of cases. 
Americans insist upon being supplied with much more water 
per capita than is usually furnished in Europe, but they often 
are singularly indifferent as to its quality. It would be a reform 
of great moment if they could be induced to curtail the present 
enormous waste of public water, such as that of one New York 
city for instance, which is 70 per cent of the entire pumpage, f 
and to expend the money thus permitted to leak away in a 
vigorous effort to improve the quality of the supply. No such 
lowering of the typhoid death-rate as occurred at Munich, San 
Remo, and sundry other places could be looked for, perhaps, 
but a large percentage of the present rate could be cut off, and we 
think from a consideration of the above figures that such a re- 
duction would pay. 
No weight should be attached to the argument, so often 
advanced by the individual householder, that he and his 
family 11 have used the water without evil result for fifty 
years." A single family is too small a collection of units 
upon which to base any proper estimate touching the question 
at issue. 
Placing the former typhoid death-rate for Albany, as above, 
at seventy-five annually, it would call for one death in a family 
of five persons every 261 years, a period much beyond the 
limits of ordinary family record. 
This " argument of experience " is frequently used to de- 
fend well-waters whose purity has been questioned by the in- 
spector. It must be remembered that a well which is polluted 
by " normal " sewage may furnish, during long periods of time, 
* Vennen v. New Dells Lumber Co., Oct. 26, 1915. 
f J. N. E. Water-works Asso., xiv., 212. 100 
WATER-SUPPLY 
a water which is not disease-producing; and yet it may sud- 
denly become very dangerous because of the. introduction of 
infected sewage derived from a newly developed, and possibly 
imported, case of typhoid fever, or from the visit of some 
chronic " typhoid carrier." 
A 11 typhoid carrier " is a person who, although apparently 
in normal health, nevertheless discharges bacilli of the disease 
in stools or urine or both, and who may continue so to do for 
a period of months or even years. Such a person acts as a 
sort of reservoir for the germs and is capable of widespread 
distribution of them. In fact it is now common to attribute 
to the action of 11 carriers " those instances of sporadic, obscure 
outbreaks of the fever which were formerly always cited as 
indicating a development of the B. typhosus from a non-typhoid 
parentage. 
Data show that women are more prone to become " carriers " 
than are men in the ratio of five to one, an unfortunate fact 
in view of women being more closely connected with the prepa- 
ration of food. 
Of those who recover from typhoid fever about 3 to 4 per 
cent become 11 carriers " and remain so for varying and some- 
times for long periods of time. 
Out of 6708 typhoid patients examined in France, 310 
became " carriers " for the following lengths of time:* 
From 2| to 3 mos 144 or 2.15% 
" 3 mos. to 1 year 64 
11 1 year to 3 years 87 
" 3 years to 3I years 15 
2-47% 
310 4-62% 
The above periods of retention of the typhoid germ are 
occasionally greatly exceeded: Thus a case is reported of a 
woman who carried the bacilli for fifty-two years. Her general 
health was good except for more or less diarrhoea. In her 
* Revue d'Hygiene Municipale, 6, 338. DRINKING-WATER AND DISEASE 
101 
capacity of boarding-house keeper, during the years 1905 to 
1908, she gave the disease (presumably) to 159 of her boarders.* 
It will be remembered that Mary Mallon (" Typhoid Mary ") 
was discovered by Soper in 1907. He secured evidence showing 
that this woman had, by her working as a cook, caused 28 
cases of typhoid fever. "She was forcibly removed to a hos- 
pital, where an examination of her urine and faeces confirmed 
the suspicion that she was a typhoid carrier." 
"Subsequently, in 1910, Mary was released on her promise 
not again to hire out as a cook and to use all precautions to 
prevent the occurrence of typhoid infections. She soon broke 
her parole, and at once the results were disastrous; two cases 
of typhoid fever occurred in a private sanatorium where she 
was employed. She went to the Sloane Hospital in October, 
1914, and there followed in January and February an out- 
break of twenty-five cases of typhoid fever among the 
physicians and nurses attached to the institution. 
"The history of 'Typhoid Mary' gives point to the new 
sections of the Sanitary Code which forbid persons ill with 
infectious diseases to engage in the handling of foods sold to 
the public. It also emphasizes the importance of personal 
cleanliness, for,'disgusting as it is, there is no doubt as to the 
probable path of infection. 
" This dangerous carrier was again apprehended and de- 
tained by the Department of Health at Riverside Hospital on 
North Brother Island." f 
"Doerr cites cases reported by Drober and Hunner, in 
which the bacilli were isolated from the gall bladder seventeen 
and twenty years after recovery, and Lentz asserts that if 
after ten weeks from convalescence the excretion of the bacilli 
has not ceased, it will most likely continue permanently and 
uninterruptedly, in spite of medication. He cites a number 
of cases in which ten,, thirty, and even forty-two years after 
recovery the excretion continued." J 
* Lancet, 195, 492. 
t Bui. Dept. Health, N. Y. City, April, 1915. 
t N. Y. State Board Health Bui., June, 1907. 102 
WATER-SUPPLY 
"Carriers" are more dangerous in the country than in 
the city because of the greater chance of excreta getting in 
contact with food under country conditions. The enormous 
number of typhoid bacilli that may be discharged by a "car- 
rier" has been determined by several observers. Petruschky 
found the urine to contain 172 million per cubic centimetre; and 
Gartner reports hundreds of thousands per gram of faecal matter. 
Although foreign to our topic, it may be well to state here 
that disease " carriers" are not limited in their activities 
to spreading typhoid fever alone. Sundry other diseases, 
notably diphtheria and cholera, are also distributed through 
such agencies. Nevertheless, the typhoid "carriers" command 
our closest attention and, until we can secure light upon some 
surely curative method of treating such persons, it would scarcely 
be a harsh order to prohibit their occupying themselves with 
the preparation of food. In addition to such an order, in- 
struction should be given them, setting forth the danger they 
are to the health of the community and showing them how 
necessary it is that they should adopt strict cleanliness in 
their personal habits. 
On March 31, 1916, a bill was introduced in the New York 
State Senate to provide for typhoid carriers, as follows: 
"Whenever an individual is declared by the state commis- 
sioner of health as being a carrier of typhoid fever bacilli and 
whenever, for the protection of the public health, the state 
commissioner of health shall have certified to the necessity 
of continued quarantine; or, whenever, in accordance with 
rules and regulations adopted by the state commissioner of 
health a carrier of the germs of typhoid fever is prevented from 
carrying on any occupation which would enable him to gain 
a livelihood, such individual may be given hospital or insti- 
tional care under the surveillance of the local health officer 
at the expense of the state if such hospital or institution in the 
judgment of the state commissioner of health be properly 
equipped for the care and maintenance of said individual. 
"When no such hospital or institution is available and when 
in the opinion of the state commissioner of health such indi- DRINKING-WATER AND DISEASE 
103 
vidual may be cared for at home or in a private family with 
due regard to the protection of the public health, the local 
charities commissioner or overseer of the poor shall, in accord- 
ance with rules and regulations adopted by the commissioner 
of health, furnish necessary medical attendance and main- 
tenance." 
Houston * considers the typhoid carrier as the great agent 
in the spread of the fever; he even terms such a person as the 
"home of the typhoid bacillus" and the "factory of the disease." 
Concentrated doses of pathogenic bacteria are what we 
have to fear, for the reason that they overwhelm and break 
down nature's defence; and such heavy doses are, according 
to Houston, just what the "carriers" supply in contrast with 
what is furnished by the more "diluted" general sewage. 
He, in short, fears the unknown visitor more than he does 
the local community. 
"In a large town there can hardly be any disproportionate 
number of 'carriers,' whereas when we are dealing with one, 
or only a few individuals of unknown health history the in- 
clusion of a single carrier may enormously affect the disease- 
producing power of their combined excreta." 
It is interesting to note that he believes one function of 
a municipal filter is that of breaking up collections of disease 
germs and causing separation of the individuals, thereby pro- 
tecting the consumers from "concentrated doses" even if the 
germs should actually pass the filter. 
It must be noted that there are some typhoid "carriers" 
who have no typhoid history. They have apparently never 
had the disease, and their tolerance of the B. typhosus has 
doubtless been secured by the repeated invasion of attenuated 
typhoid germs. They have slowly acquired a partial immu- 
nity against the disease by such an invasion, long continued. 
In this connection we note the difference between the suscep- 
tibility exhibited by persons long accustomed to the use of a 
polluted water supply and that shown by the non-acclimated 
stranger. In like manner we probably become, as we grow 
* 9th Research Report Metro. Water Board. 104 
WATER-SUPPLY 
older, gradually immunised against those "children's diseases" 
to which the young are so susceptible. 
As has been already said, we should remember that when 
studying the extension of typhoid fever, or for that matter, 
disease in general, the power of the natural bodily resistance 
must be ever kept under consideration. 
Certain invading organisms, that of plague, for instance, 
are so virulent in their attack that they are well nigh certain 
to break down nature's barrier of defence, and the lodgment 
of the germ becomes synonymous with the taking of the dis- 
ease. Others, like that producing pneumonia, are very often 
carried about in the bodies of well people and require a decided 
lowering of "the power of resistance" before their pathogenic 
activities can become effective. Typhoid fever occupies a 
position between these extremes. Its invasion is a serious 
matter if its germs be vigorous and the "resistance of the 
subject" be low; but if the bacilli be attenuated or the resisting 
power of the subject be high, then the disease does not ma- 
terialize at all, or the case developed is a mild one. Accepting 
it to be true that a disease is caused by the activities of a 
specific germ and that in the absence of the germ such disease 
cannot occur, and accepting it to be also true that a lowering 
of "resisting power" is frequently followed by outbreak of 
disease, then it follows that "unsuccessful invasion" must 
often take place. We are attacked and resist the attack more 
frequently than we appreciate. 
An apparent increase in susceptibility is often observed 
during those typhoid epidemics which originate in the pollu- 
tion of a city's water supply by the city's own sewage. In 
such cases the disease feeds itself, the sewage, and consequently 
the water supply, becoming progressively more infected. 
Diarrhoeal outbreaks of more or less intensity are frequently 
observed to precede epidemics of real typhoid, and they serve 
as warning flags of danger to come. 
Having a shorter incubation period the milder disease develops 
first, and fewer people appear to be immune against its attack. DRINKING-WATER AND DISEASE 
105 
Sewage pollution of the water supply is commonly the 
reason for such disturbances of the public health, and it may be 
that typhoid will not follow if the cause of the trouble be re- 
moved in time. 
Dr. Ball, of Warren, Pa., reports an outbreak of gastro- 
enteritis in 1906 which was unquestionably due to an infection 
of the water from a group of wells driven on an island in the 
Alleghany River. Suddenly, with no warning, some 1800 
cases of the disease developed. There were no fatalities, but 
the cases were severe, often approaching collapse, and the 
illness lasted in most instances approximately thirty-six hours. 
The incubation period was about one day. Upon investigation 
there was found to be quite a direct connection between one 
or more of the wells and neighboring drains. There was no 
subsequent typhoid, but of course the source of the public 
water supply was changed. 
Sedgwick reports a similar epidemic of diarrhoea at Bur- 
lington, Vt., from sewage pollution of the lake supply. 
Thresh * gives an instance of a serious outbreak of diarrhoea 
caused by water from a service tank which was found to con- 
tain "many worms and other objectionable forms of small 
animal life." 
It is an unfortunate fact that these diarrhoeal troubles do 
not, in most instances, terminate so fortunately as in the cases 
just given, it being more common for typhoid fever to follow 
hastily upon the heels of its less fatal forerunner. 
Emergency intakes may be the means through which ill- 
ness and death are spread broadcast throughout the com- 
munity. The underwriters very naturally insist upon a sufficient 
fire service, which shall be available in the event of a tem- 
porary breakdown of the regular distribution system. It too 
often happens that upon such occasions a very inferior water 
is supplied by the "emergency intake." and as a result of its 
use there follows an outbreak of typhoid fever. Commonly, 
some old intake is allowed to remain in place for "emergency 
* Exam, of Water, page 46. 106 
WATER-SUPPLY 
service" when pollution of the former supply has so grown 
in intensity as to force the authorities to seek a new source 
for public water. 
Further fouling of this old supply goes on progressively 
as population increases, until after some years the water be- 
comes practically dilute sewage. Suddenly an accident to 
the regular water system induces the authorities to open the 
old gates, and the result may be imagined. 
/ 5 Cast Iron Intake Pipe"^- 
Cascade Creekl/ Water Works 
„ . , Pumping Station X 
Little Cascade Creek f 
CITY OF ERIE {Mill Creek 
AT ERIE, PA., POLLUTED WATER ENTERED THE PUMP-WELL THROUGH A 
LEAK AROUND THE OLD INTAKE GATE NEAR THE PUMPING STATION. THE 
RESULT WAS A MOST SERIOUS TYPHOID EPIDEMIC. 
Such has been the history of typhoid epidemics in a num- 
ber of cities. 
Thus at Butler, Pa., on October 20, 1903, the filtered water 
supply became temporarily unavailable and polluted water 
was admitted to the mains from a former river intake. A 
serious epidemic of typhoid fever followed, resulting in 1270 
cases and 56 deaths. 
Storage for a sufficient length of time, supplemented, 
if necessary, by an appropriate dose of bleaching powder, 
will render even a poor water acceptable for emergency uses, DRINKING-WATER AND DISEASE 
107 
and the reservoir capacity for such storage need not be 
large. 
Abandoned intakes should be permanently cut off and water 
for emergency use should be drawn from a source the safety of 
which can be assured. 
Vacation typhoid is a term used by not a few physicians 
to represent those cases of the disease which appear in the 
early autumn and which have been engendered by use of pol- 
luted water from badly located country wells, aided by sundry 
hygienic errors which are in contrast with the regularity of 
home life during the balance of the year. 
Again we are reminded of the influence of unusual environ- 
ment upon the 11 power of resistance," the same environment 
acting quite differently upon different persons. 
11 Recently studies of typhoid in India have shown that the 
maximum prevalence among the white troops is in the hot season 
as with. us, while among the natives the maximum occurs 
during the cold and rainy months. This seems conclusive 
proof that the organism of the host is affected by abnormal 
and inclement seasons." * 
Paratyphoid fever is a water-borne disease lately differen- 
tiated from typhoid fever proper. 
11 Although this is a less severe and less fatal disease than 
typhoid, it must be understood that the term is not applied to 
light cases of typhoid, but represents a distinct infection sui 
generis, probably caused by organisms intermediate between 
the typhoid and the colon bacillus." 
"The absence of diarrhoea in the great majority of cases 
is one point of possible clinical differentiation from typhoid 
fever." f 
During legal proceedings undertaken of late in connection 
with questions concerning damage to water supply, it was 
* Winslow, Am. J. Pub. Hygiene. 
t J. Infectious Diseases, Jan. 2, 1904. 108 
WATER-SUPPLY 
suggested that "anthrax" should be held accountable for a 
portion of the trouble. 
Anthrax spores are very resistant. Hastings has reported 
them to persist for eight years in pond water.* 
In man the disease is commonly produced by the infection 
of a wound, even a small break in the skin being sufficient 
for the entrance of the spores. More rarely the conveyance 
is by flies, dust, or water, but human beings are more resis- 
tant to anthrax than are most animals and the number of 
cases due to ingestion is so small as to exclude it from serious 
consideration as a water-borne disease. Thus there were but 
five cases in all England during the five years 1899-19034 
It must be admitted, however, that danger to stock does 
exist in the waters of streams that receive the industrial wastes 
from nearby tanneries, unless such trade effluents have been 
properly treated before admission to the stream. Heavy doses 
of chlorine added to the screened effluent will kill the anthrax 
spores. How much disinfectant should be used must be deter- 
mined for each case, inasmuch as the quantity of organic 
matter present will to a certainty greatly differ with change in 
local conditions and consequently widely varying amounts of 
the chlorine will be rendered inert thereby. 
New construction may temporarily greatly pollute a water 
passing through it. 
Without careful inspection the worst kind of filth may 
be deposited in the very mains themselves by the thoughtless, 
careless employees at work upon the job; and even with the 
best supervision a newly-laid pipe or a newly-finished tunnel, 
or reservoir cannot be expected to be ideal in the matter of 
cleanliness. 
Alvord reports that it was necessary at Gary, Indiana, to 
draw one hundred million gallons of water daily through the 
new. tunnel for a period of twenty-one days in order to make 
the water safe for domestic use. 
* Science, June 25th, 1909. 
t See J. Roy. San. Inst., xxvii., 698. DRINKING-WATER AND DISEASE 
109 
At Troy, N. Y., when new connections were made with 
the Tomhannock tunnel, Diven secured the same results at 
less cost by the use of heavy doses of bleaching powder. 
In view of the difficulty, so often encountered, of interpret- 
ing the terms found in many water contracts, it is well, before 
closing this chapter, to inquire 
What is a "good, pure, wholesome" water? 
Adjectives such as the above, with the occasional addition 
of the expression " clear," are commonly found in the con- 
tracts made by private water companies with the cities which 
it is their business to supply. 
Litigation over an alleged violation of such a contract 
having been instituted, it usually becomes the duty of some 
expert witness to pass upon the question as to whether or not 
the water under consideration is of such a quality as to fall 
within the limits specified by the wording of the agreement. 
To the average layman there would appear to be no am- 
biguity in describing a water as "good, pure, and wholesome," 
and he would see no objection to demanding that the words 
be descriptive of every public water supply in the country, 
and that the water companies, wherever located, be forced 
to live up to the exact and literal meaning of each and all of 
those words. 
To show, however, that the above expressions are some- 
what loosely used and that they lack in the definiteness which 
is expected of them, let us consider them separately for a 
moment. 
What is a "good" water? Surely before answering we must 
inquire: "Good for what?" If for drinking, then its considera- 
tion would properly come later when we dwell upon the ex- 
pression "wholesome." Outside of the question of potability, 
a water charged with calcium sulphate may be very "good" 
for the brewing of light ale and very "bad" for either boiler 
or laundry use. It is to the selenitic character of its water 
that Burton-on-Trent largely owes its reputation as a great 110 
WATER-SUPPLY 
ale-producing centre, for the properties which are so popular 
in the brews of Bass and Allsopp seem to depend upon the 
presence of calcium sulphate in the water from which they are 
made. Such a water would, however, be very unacceptable 
to the large laundry interest of Troy, N. Y., where the. daily 
soap-consumption is measured by the ton. 
As with "good," so with the word "pure." It is, as used, 
quite ambiguous. Strictly pure water is a chemical curiosity, 
very difficult to prepare and perhaps still harder to preserve, 
as all know who have read about Sir Humphry Davy's classic 
determination of its composition. The word in the contract 
manifestly cannot mean water such as that; but if not, then 
what does it mean? If some permissible limit of impurity 
be implied, what is that limit and what is the character of 
the permissible impurity? These are fair questions, and they 
are pressing for an answer. 
We turn now towards the expression "wholesome," trust- 
ing that here at least we have a definite term concerning which 
no misunderstanding can arise. It certainly does sustain more 
close scrutiny than the two already considered; but even this 
word, of apparently clear meaning, may be quite befogged 
by considerations that remind us of the old adage, "What is 
one man's meat is another man's poison." 
How often we note that a change from a soft to a hard 
water, or vice versa, is followed by intestinal derangements, 
particularly among young children, a change of water being often 
nearly or quite as pronounced in its results as a change of air. 
In view of this we begin to feel that Webster's definition 
of "wholesome" as "tending to promote health" may not 
strictly apply to the case in question; for it seems not un- 
likely that a shrewd cross-examining lawyer might very readily 
force the witness to admit that the water was "wholesome" 
for a part of the community only, namely, those who were 
acclimated, who had become tolerant to its use. 
What has been said regarding change in hardness applies 
with equal force to variation in turbidity. In the stranger 
unaccustomed to its use, a turbid water will often produce a 111 
DRINKING-WATER AND DISEASE 
transient form of intestinal disturbance. Shall we therefore 
declare such a water to be unwholesome? If so, our verdict 
would to a certainty be opposed by many thousands of people 
living in our great central basin. 
Appreciating the two horns of the dilemma, the writer has 
adopted this position, namely, assuming the absence of sewage 
material and considering the silty turbidity alone, a some- 
what turbid water is to be classed as undesirable for many 
purposes for which a city supply is used, but yet it cannot be 
rated as unwholesome, for the reason that it is unproductive 
of disease among the people for whose use it is primarily in- 
tended. 
As we have seen, a water suited to the uses of one town 
might not be the best for the interests of another; or perhaps 
even a portion of a city might, because of its manufacturing 
establishments, desire a different water from that sought after 
by the residential section of the same community. 
As a substitute for "good, pure and wholesome," the ex- 
pression "safe and suitable" might be with advantage inserted 
in a water supply -contract; and it is well to add here that 
the word "safe" alone will not suffice. Irrespective of its 
"safety" as regards disease production, a water should not 
offend the Eesthetic sense. Few people would knowingly care 
to drink the effluent from a sewage disposal plant, and yet some 
such effluents taste and look like spring water, and some people 
actually do drink them. It is perfectly reasonable for the 
public to demand a "suitable" raw water and to insist that, 
however perfect the process of artificial purification, the raw 
supply shall not be derived from a source unduly foul. 
A notable case in point is that developed during the "Chi- 
cago Stock Yards Water Purification Litigation." * The fol- 
lowing is extracted from Johnson's report of that case: 
"From a sanitary standpoint the whole case rested upon 
whether or not the waters of Bubbly Creek, primarily foully 
impure but purged by the purification process of all their 
visible impurities, and rendered clear and sparkling, but still 
* Engr. News, Sept. 29, 1910. 112 
WATER-SUPPLY 
containing in a dissolved state about 50% of the organic 
matter which the raw water contained, in the practical ab- 
sence of bacterial life, and in the absence so far as bacteriological 
tests now can show of all bacteria commonly known as disease- 
producing forms, could still be considered a satisfactory water 
with which to water live stock. 
"This filter plant was not constructed for the purpose of 
producing from Bubbly Creek a water supply .for consump- 
tion by human beings. For some reason or other, as the trial 
progressed, the city pushed this phase of the question more 
and more to the front, until ultimately it was made to appear 
that, among other things, the company had installed this 
plant to render the waters of Bubbly Creek safe for human 
consumption, and that one of the final objects the Stock Yard 
Co. had in view was to substitute for the city water the effluent 
of the filter plant to be used throughout the yards for human 
consumption, as well as for watering live stock. 
"Is water so badly polluted as that of Bubbly Creek a 
proper source of drinking water? 
"This is a much-mooted question among sanitarians, but 
one which is far from being settled. Where the line may 
be drawn between polluted waters which may be rendered 
safe by filtration and sterilization processes and those which 
may not is beyond the writer's ability to indicate at this time. 
"It is true that the waters of Bubbly Creek are much fouler 
than those ordinarily encountered in problems of water puri- 
fication. The difference between the purification of such 
water and waters of far greater purity is, from a broad, prac- 
tical standpoint, only a matter of degree. 
"As to the effect on health of organic matter and its de- 
composition products in drinking water, the city endeavored 
to bring out the fact that, even though by the purification 
process the raw water was made pleasing to the eye and rendered 
practically free from bacterial life, the filtered product still 
contained dissolved organic matter in such quantities as would 
be likely to exert a toxic influence on persons drinking such 
water continuously. DRINKING-WATER AND DISEASE 
113 
"Anticipating that this contention would be raised the 
experts for the Stock Yards Co. considered carefully all of the 
available evidence bearing on this question. As a result the 
conclusion was reached, which conclusion was substantiated 
by the writings of numerous authorities on the subject, that 
in the absence of bacterial life organic matter is not injurious 
to the public health when consumed in such quantities as were 
found in the filtered water at the Union Stock Yards. 
"One important witness testified in connection with this 
question that: 
" 'From my experiments I concluded that substances which 
are to be considered harmful are not present in the filtered 
water.' 
"In his cross-examination this point was again brought 
up as follows: 
" 'Q.-What limit would you consider as the amount of 
organic matter that might be in water which was practically 
sterile so far as bacteria are concerned? 
" 'A.-I could not give any exact limit, but undoubtedly 
a larger amount of organic matter than we have in this case 
would still be within a reasonable limit.' 
" 'Q.-Do you consider that organic matter that comes 
from sewage is not more harmful than water which has organic 
matter in it that comes from decayed vegetables?' 
" 'A.-When you are talking about water purification, there 
is absolutely no difference in it. The protein matters are the 
same. It is a well-established fact that they yield the same 
decomposition products. 
"Attention is also directed to the evidence given by another 
witness, who in this connection testified on cross-examination 
in this case as follows: 
" ' Q.-If you were to select a water supply for a community 
and have the two sources of supply, Bubbly Creek, with this 
filtration plant as operated, and Lake Michigan, with the water 
that we could get from Lake Michigan, in the shape that it is 
in, would you recommend the Bubbly Creek supply instead 
of the Lake Michigan supply?' 114 
WATER-SUPPLY 
" 'A.-I should say that it would be more safe and usable 
than the other. Yes.' " 
Inasmuch as the people usually know whence it is pro- 
posed to take the water, would it not be wisdom to place in 
the contract some form of definite analytical specification, where- 
by the water, as offered to the consumer, could be held up to 
the quality established by an accepted local standard. 
In conclusion, let a word be said about a matter that has 
recently awakened no small degree of general interest. 
It would seem that the attention of the public has been 
largely drawn of late to the question of the wholesomeness of 
distilled water for dietetic purposes, numerous articles having 
issued from the press under such captions as "Poisonously 
Pure Water" and the like. 
Much uneasiness has been created among those who have 
substituted the use of distilled water for domestic purposes, in 
place of some supply of acknowledged impurity, because of 
their being told that " distilled water is an active protoplasmic 
poison, due to its property of extracting salts from animal 
tissues and causing them to swell up by imbibition." 
Stress is, of course, laid upon the increased danger arising 
from the use of such water by the young, whose tissues are in 
process of formation. 
Surely it is only fair to insist that the burden of statistical 
proof be borne by those who advance this proposition, and 
until such data are furnished we will rest our faith upon such 
facts as the following: 
The Croton water supply of New York City contains 5.48 
grains of solid matter per U. S. gallon, or 1.37 grains per 
quart. 
One quart of water per day for a child would be a liberal 
allowance, considering that milk is also a portion of the diet. 
In consideration of the mixed character of human food, 
and the amount of mineral salts naturally occurring therein 
and added thereto, the withdrawal of 1.37 grains of mineral 
salts per day, by the substitution of distilled water for that 115 
DRINKING-WATER AND DISEASE 
drawn from the Croton River, would be a matter of too small 
importance for consideration. 
Finally, let it be remembered that all the vessels of our 
navy condense their drinking-water by means of the Baird 
condenser, and the navy medical authorities report most ex- 
cellent results. Surely the bone and muscle of the "men 
behind our guns" speak for themselves. 
In a letter to the author, Surgeon-General W. C. Braisted 
of the U. S. Navy says: "The use of distilled water on ships 
of the Navy has always tended to the very best of health 
conditions. In my opinion the use of distilled water offers 
the ideal drinking fluid for human consumption." 
It may be interesting to add that "the Marine battalion, 
some five or six hundred strong, used distilled water from the 
ships all the time they were doing shore duty at Guantanamo, 
Cuba, and they had none of the enteric fever that prevailed 
so extensively in the Fifth Army Corps." CHAPTER III 
ARTIFICIAL PURIFICATION OF WATER 
To comprehensively cover so broad a subject as is indi- 
cated by the heading of this chapter would call for a series of 
monographs rather than the short space that can be here 
devoted to it. Water purification has become a special art 
that has been again divided into sub-specialties, each with its 
own experts and an extensive literature. 
It must be accepted as an axiom that pure water is better 
than purified water, but pure water is becoming more and 
more difficult for many of our towns to secure, so that the best 
that some of them can hope to obtain is a polluted water which 
has been efficiently purified by art. It is a question, some- 
times, whether it would not be better policy, considering the 
rapid changes in the density of population, to accept a 
moderately polluted source and thoroughly purify its water, 
rather than go to large expense in obtaining a faultless 
supply which might have to be purified in its turn at some later 
day. 
It is hopeless and unreasonable to expect an up-stream 
community to so purify its sewage and otherwise protect the 
water of a stream which it has a right to use, as to deliver its 
water to the down-stream neighbor 11 undiminished in quantity 
and undamaged in quality." The common law may order 
such neighborly action, but the demand is both impracticable 
and unfair. Inasmuch as the lower city should, in any event, 
erect purification works to care for its water supply, it would 
be unjust to require those living higher up to attempt the 
conversion of their sewage and trade-waste into drinking water 
before turning the effluent into the stream. The burden of im- 
provement should be divided, and all that could in fairness be 
116 ARTIFICIAL PURIFICATION OF WATER 
117 
asked should not go beyond so protecting the water of the 
stream as to permit of its being delivered to the water-works 
intake in such condition as to allow of its being used without 
undue difficulty or expense, as a "raw" water for modern puri- 
fication works. 
And that brings up the question, How bad must a water 
be to require its exclusion as being unfit for a "raw" supply? 
No definite reply can be made. If necessity demanded it, and 
expense were not considered, a water safe for drinking purposes 
could doubtless be prepared from sewage itself. It is difficult 
to conceive of a raw supply much worse than that of Bubbly 
Creek, Chicago, and yet an attractive water is the output 
of the purification plant used to improve it. 
The poorer the raw supply the greater is the risk of trouble 
following careless handling or temporary inefficiency of the 
purification works. A sensible rule to follow is to select 
the very best source bf supply available and then improve 
the same, if necessary, by the most reliable process suited 
to the needs of the case. 
Outside of any question of wholesomeness, the water 
which a part of our people are content to use at present may 
not be considered suitable in a few years to come. The ten- 
dency is towards a general demand for a clear and colorless 
water, and water-purveyors must be prepared to meet it. 
Filtration of surface water, before delivering the same for 
public consumption, is now specifically ordered by the laws 
of Germany, and rules are laid down for its proper accom- 
plishment. Such legislation is not improbable in this country, 
thus still further making it the part of wisdom to anticipate the 
artificial improvement of some waters which are now possibly 
considered beyond impeachment. 
The art of removing suspended material from water by 
some form of filtration has been known during many ages, 
although it was not put into extended practice until very recent 
times; thus the siphoning of liquid from one vessel to another 
by the capillary action of porous material, such as a strip of 
cloth, and the consequent separation of the liquid from sus- 118 
WATER-SUPPLY 
pended material, was well known to the ancients and is fre- 
quently mentioned.* 
The modern methods of filtration claim to do something 
more and better than merely to strain off the grosser elements 
of turbidity; and so fully do the people of Europe appear to 
believe this claim a just one that with them a city water-works 
without an attendant filter-plant is becoming almost a novelty. 
The method of purifying water on the large scale which de- 
serves first attention on account of its early use and wide 
application is that of "slow sand filtration," commonly known 
also as 
THE ENGLISH FILTER-BED SYSTEM 
Briefly described, an English filter-bed is a tight reservoir, 
containing some five or six feet of stratified filtering material, 
of progressive degrees of fineness, beginning at the bottom with 
broken stone or gravel, and ending with an upper layer of fine 
sand, the whole being suitably underdrained. 
Much diversity exists in the relative thickness of the several 
layers, some filters being constructed with a very thick upper 
layer of fine sand, while with others the finest material is put 
on as a comparatively thin cover. The Dutch filters are 
especially marked in the thinness of their beds,f a feature by 
no means to be recommended; for, although much of the 
actual work of filtration is done by the upper layer of sand, 
yet, if the thickness of the body of the bed be unduly reduced, 
that portion of the water which is in the act of being delivered 
will bear too large a ratio to that filling the interstices of the 
coarser layers; as a result, currents will be established and 
ruinous channelways be quickly worn in the uppermost stratum. 
Through the fine-sand layer the water slowly and evenly 
passes, leaving the bulk of its suspended impurities upon the sur- 
face to form theSchmutzdecke, or "dirt cover," of the Germans. 
Beyond the mere gradual accumulation of suspended matter 
strained from the water, this Schmutzdecke is in part composed 
* Bolton, " Ancient Methods of Filtration," Pop. Sci. Monthly, xvi., 495. 
f Mr. Halbertsma informs the author that the height of the ground-water level 
is one reason for the thinness of the beds. ARTIFICIAL PURIFICATION OF WATER 
119 
of slimy, jelly-like material, produced through bacterial agency, 
which serves to entangle and hold bacteria and other sus- 
pended substances of all kinds. 
SECTION OF, AN ENGLISH FILTER-BED. SOUTHWARK & VAUXHALL WATER CO., 
LONDON. 
Fine 
Sand. 
Coarse 
Sand. 
Fine 
Gravel. 
Medium 
Gravel. 
Coarse 
Gravel. 
Small 
Stones. 
Large 
Stones. 
Total 
Depth. 
Berlin 
22 
2 
6 
5 
3 
4 
12 
54 
Zurich 
32 
6 
4 
6 
48 
London, Southwark 
and Vauxhall 
36 
12 
9 
9 
66 
Poughkeepsie, IN. Y.. 
24 
18 
6 
24 
72 
Hamburg 
40 
24 
64 
Albany, N. Y 
48 
22 
2 
6 
584 
Roxborough,Phil.,Pa. 
36 
I 
5 
4 
6 
52 
Pittsburgh, Pa 
48 
15 
2 
31 
5 
60 
COMPOSITION OF VARIOUS FILTER-BEDS, IN INCHES 120 
WATER-SUPPLY 
It must not be thought, however, that the extreme top 
layer of sand, with its cover of slime, does the entire work, so 
far as purification is concerned. Each of the sand grains of 
the body of the bed becomes covered with a sticky coating of 
the zooglcea jelly, and they collectively are to be credited 
with a large share of the results accomplished. This is shown 
by Reinsch, who has published his observations of the Altona 
filters. He found the unfiltered water to contain 36,320 bac- 
teria per cubic centimetre. After passing the slime layer 
there yet remained 1876, but after going through the entire 
depth of sand there were found but 44 per cubic centimetre. 
Thus the lower layers have uses other than mere regulation of 
flow. 
A filter recently disturbed by the process of cleaning is 
lower in efficiency than one which has been for some time 
in operation. In fact it might be properly described as "at 
its best when at its dirtiest"; and, were it not that it finally 
becomes almost impervious to water, it would be better not 
to clean it at all. A new filter is of small use until it 
"ripens" for work; that is, until the "nitrifying" organisms 
have firmly established themselves and the zooglcea jelly 
envelops the sand grains. We have here an additional reason 
for a thick sand layer which will permit more of the filter to 
rest undisturbed. 
The thickness of the sand layer, for proper working, should 
be made not less than thirty-six inches, and this depth should 
not be permitted to greatly decrease, by reason of the suc- 
cessive removals of layers of the upper surface for purposes 
of cleaning. At Albany the thickness of the fine-sand layer 
is not allowed to fall below the limit of two feet. 
The head lost in passing the filter naturally increases as the 
sand becomes clogged, and the rapidity with which this loss 
occurs will of course vary with the character both of the bed 
and of the water being filtered. When this loss of head reaches 
four feet at the Albany plant the filter is shut down for clean- 
ing. Such figure is a common one. 121 
ARTIFICIAL PURIFICATION OF WATER 
The engineering structures containing these layers of filter- 
ing materials differ from one another in size, in shape, and in 
method of construction, according as the preference of the 
designer may dictate or the necessities of the case may demand. 
In London it has been common to select one acre as the 
proper superficial area of a filter-bed. The Hamburg filters 
are each 1.89 acres in area; those constructed at Albany, 
N. Y., are each seven-tenths of an acre, while the areas 
of the individual beds at Philadelphia (Roxborough) and 
Washington are one-half and one acre respectively. Usually 
the inner wall-surface is nearly, or quite, vertical, but the 
Holland filters (see page 122) form a notable exception in 
this particular, having a slope, at times, of more than one to 
one. 
An objection to an entirely vertical wall is that there is 
possibility of improperly filtered water passing down between 
it and the sand. A wall having some slope or one broken 
into steps affords a better opportunity for a good joint being 
made with the filtering material. 
As to composition, the body of the side walls is as varied 
as one would expect to find it among reservoirs in general, 
running from earth embankments with clay puddle cores to 
structures of pure concrete, or even of dressed stone. Such as 
are constructed of earth are, however, carefully protected 
on the inside, by suitable paving, from the damaging action of 
ice and waves. 
Some of the filters of the Southwark and Vauxhall Com- 
pany (London) have a layer of three-inch agricultural drain- 
pipe, placed side by side with open joints, over the entire 
bottom, thus securing easy flow to the clear-water reservoir. 
(See illustration, page 119.) 
In the Albany plant these are replaced by lines of six-inch 
vitrified pipe, open-jointed, and laid thirteen feet eight inches 
apart. 
A convenient form of hollow-tile underdrain has been in- 
vented by L. K. Davis. It is easily understood by reference 
to the illustration on page 125. 122 
' !5?7 FineSand 
• 7. 9 Coarse n 
7:'9 Fme brave! 
.Ui'dCoarse »> 
WATER-SUPPLY 
l5'.'7Sand 
9?8Fine braved 
S."9 Coarse n 
27n5Sand 
n"d brave! 
Primary, ।Secondary 
Filter. J ' 
Zuetphen. 
Utrecht 
Secondary Filter., 
• 13 QFmeSand 
K 2 " Coarse •• 
r 4" Sea Shells 
II?8 Coarse , 
^'H'8 brave! 
5 BrukenSrone 
24'- 30"Sand 
- S "9 Coarse 5and 
% 2 ."9 Fme brave! 
9:'8 Coarse •» 
IT 19?7 Coarse Sand 
9.8 Fme brave! 
141 2 Coarse »» 
Schiedam. 
i9~7Sand San^BJ' 
lireSeaSMIs 
7"8 Coarse 9°™%?. 
brave! brave/. 
Middelbur^. 
(After Halbertsma.) 
Vechtwasser. 
Primary Filter. 
Groeningen. 
- 7 8 Coarse Sand 
- 7 '8 Fine »> 
ll "8 Coarse t 
7 "0 Fme brave! 
Il"6 Coarse •• 
ri. 
J 
24 -30 "Sand 
7 5 "9 Sea Shells 
* 2 "9 Fine bravel 
918 Coarse m 
Amsterdam. 
H'. 8 Coarse5and 
7'. '9 Fine n 
9?8 Coarse 
9"8 brave! 
19'.'7 Broken Stone 
Sliedrecht, Gonrichem. 
Enschede. 
SECTIONS OF DUTCH FILTER-BEDS. 
s Duenenwasser. 
■ 7 8 Fine Sand 
' 7" 8 Coarse „ 
< 3"9 Sea Shells 
Z- 3 " 9 Fine brave/ 
* 7 '8 Coarse n 
Dordrecht 
!9'.'7Sand 
!^"7 Sea Shells 
Leeuwarden 
7 u"eSand 
T 9"8SeaShe!!s 
I 9 "8 brave! 
r s:'9 Broken Smne 
\ 9"8 Fine Sand 
■■ !7 "7Coarse -* 
7 9 "eurave! 
| ^Courses ofBnch 
Leiden.. 
!9'.7Sand 
77 Fine brave! 
!2"6Coarse n 
Ha^ue. 
Vlaardin^en. 
Rotterdam. ARTIFICIAL PURIFICATION OF WATER 
123 
Covers for filter plants add greatly to the expense and, in 
mild climates, the increased cost may not be warranted, but 
in those localities where the winters are severe it becomes 
necessary to throw over them a cover, which is commonly 
of concrete, resting upon columns of the same material. 
Thick ice renders it practically impossible to properly clean 
a filter by the ordinary methods, and the resulting imperfect 
purification of the filtrate is often coincident with increase in 
the death-rate. This was noted in Berlin in the winter of 
1889, when an outbreak of typhoid fever followed the deficiency 
in purifying power of the open filters. That portion of the city 
supplied with water from the covered filters was not visited 
by the epidemic. 
Protection against ice is not the only advantage offered 
by the cover, however, Heavy growths of algae may occur in 
the water of an open filter, with serious clogging of the sand. 
The writer has seen filters in England so blocked by growths 
of Mougiotia that until the matted masses were removed prac- 
tically no water could pass the sand. Exclusion of sunlight 
prevents such growths. 
Increased difficulty of cleaning is not worthy of consid- 
eration as an objection to covers, for with modern methods 
both open and covered filters are cleaned with equal facility. 
When it is a doubtful question whether or not covers will 
be necessary, it is good policy to put in the foundations for the 
piers necessary for their support so that later the covers, if 
required, may be added without undue expense. 
In England the climate does not demand the construction 
of covered filters as protection against frost, as much trouble 
from ice is not experienced; but even there exceedingly cold 
weather will at times occur, bringing with it large additions 
to the bill for expenses of maintenance. A notable winter in 
this particular was that of 1884, when seventy men were con- 
stantly employed in removing ice from the Southwark and 
Vauxhall beds. (See illustration, page 124.) 
Mr. Allen Hazen, in his excellent work on "Filtration of 
Public Water-supplies," advocates the covering of filters in 124 
WATER-SUPPLY 
REMOVAL OF ICE FROM LONDON FILTER-BEDS. ARTIFICIAL PURIFICATION OF WATER 
125 
all localities where the mean January temperature is below 
the freezing-point. 
An idea of the interior appearance of a covered filter may 
perforated tile underdrains. (Devised by L. K. Davis.) 
be obtained from the cut on page 129 showing a filter belong- 
ing to the plant at Ashland, Wis. 
The illustraton on page 130 is from a photograph showing 126 
WATER-SUPPLY 
GENERAL PLAN 
SCALE OF FEET 
0 100 
0 10 20 30 40 50 
SCALE OF METERS 
SEDIMENTATION BASIN 
382's"x 600'0" 
FILTERS AT ALBANY, N. Y. (Hazen.) ARTIFICIAL PURIFIICATON OF WATER 
127 
FILTER BEDS 
SECTIONS OF FILTERS 
SCALE OF FEET 
0 1 2345 678 
_ 1 
SCALE OF METERS 
Albany, N; Y., filter-beds. (After Hazen.) 128 
WATER-SUPPLY 
the covered beds at Zurich, Switzerland, in process of con- 
struction. 
As is well known, London filters all of its water, with the 
exception of what is derived from deep wells in the chalk. 
Certain statistics relating to these plants are here given.* 
Name of 
Company. 
Source of Supply. 
Daily Supply 
in 
U. S. Gallons. 
Number 
of Filter- 
beds. 
Total 
Area of 
Filter- 
beds, 
in Acres. 
Depth of 
Filter 
materials. 
Chelsea 
Thames river 
12,727,000 
7 
8 ft. 
Thames river 
Lee river 
East London.. . 
Chalk wells 
51,495,000 
31 
291 
3 ft. 6 in. 
Springs J 
Grand Junction 
Thames river 
22,391,000 
15 
5 ft. 6 in. 
Kent 
Deep Chalk wells.. . 
17,126,000 
none 
Lambeth 
Thames river 
23,509,000 
IO 
9i 
7 ft. 
f Chadwell sprirg 1 
New River.... 
1 Lee river. > 
43,190,000 
20 
i6j 
5 ft. 7 in. 
I Chalk wells J 
Southwark and 
Vauxhall. . . . 
Thames river 
34,080,000 
12 
. i4i 
5 ft. 6 in. 
West Middlesex 
Thames river 
21,627,000 
12 
15 
5 ft. 6 in. 
Total. . . 
226,145,000 
107 
1093 
The cost of constructing a filter-bed upon the general plan 
described must necessarily greatly vary, in direct ratio, with 
the local cost of labor and materials, and with the difficulty of 
the engineering problem involved. 
For some well-known plants the cost of construction is 
given, as follows, exclusive of the price of land: 
London.-A bed one acre in filtering area costs from $24,000 
to $39,000, depending on the nature of the ground. Those 
of the Southwark and Vauxhall plant each cost the latter 
sum. All these beds are uncovered. 
Liverpool.-Same as London. 
Zurich.-Covered beds, complete, cost 120 francs per square 
metre of filtering surface (about $2.25 per square foot, or 
* For further information see " The London Water-supply," by Shadwell. WATER-SUPPLY 
129 
$98,000 per acre). The uncovered beds, previously in use, 
cost two-thirds of this sum. 
Hamburg.-The filters when first built (open), cost 33 
marks per square metre of sand-surface (about 70 cents per 
square foot, or $30,500 per acre). 
A COVERED FILTER AT ASHLAND, WIS. 
Berlin.-The covered filters cost $70,000 per acre, and 
the open ones about two-thirds that sum. Lindley gives a 
general estimate for the Continental filters, as follows: Open, 
$45,000 per acre; covered, $68,000.* 
* See also Engineering News, Aug. 16, 1894. 130 
WATER-SUPPLY 
COVERED FILTER-BEDS AT ZURICH, SWITZERLAND, UNDER CONSTRUCTION. ARTIFICIAL PURIFICATION OF WATER 
131 
Ashland, Wis.-The beds are three in number; they are 
covered and of one-sixth of an acre each. The complete cost 
was at the rate of $70,000 per acre.* 
Albany, N. Y.-The filters are eight in number, of con- 
crete, with vaulted covers, also of concrete. Cost per acre, 
$45,600. This does not include the pre-filters or "scrubbers" 
since added. Miscellaneous data concerning these beds will 
be found on page 144. 
Mr. G. W. Fuller gives the following (complete) costs per 
acre: 
Augusta, Me $94,534 
Pittsburgh, Pa 72,777 
Philadelphia (Queen's Lane) 71,550 
Washington, D. C 74,000 
An approximate idea of the cost of some older filter-beds 
may be obtained from the estimates given by Mr. Edmund B. 
Weston in a paper in the Engineering News of October 7, 1897. 
Covered with masonry vaulting, cost per 
acre $70,000 
Uncovered, cost per acre 38,892 
These estimates do not include the cost of a sedimentation- 
basin. 
In Mr. Hazen's opinion covered filter-beds should be roughly 
estimated as costing $15,000 per acre more than those which 
are open.f 
Sedimentation-basins are essential in all cases where the 
water to be filtered is materially turbid. These basins need 
not be of great size. Storage sufficient to equal the twenty- 
four hours' supply is quite enough, for in that time the great 
bulk of the suspended material will settle, and the balance can 
be economically removed by the filter. Where no settlement 
is permitted before running a turbid water upon the filter, 
basin. 
* J. N. E. Water-works Asso., xi., 314. 
f Engineering News, Jan. 24, 1902. 132 
WATER-SUPPLY 
an unnecessarily rapid clogging of the sand results, with con- 
sequent increase in frequency of cleanings. 
A clear-water basin is required to receive the filtrate from 
a filter plant, in order to provide for fire emergency. One 
capable of holding eight hours' supply is sufficient. Under 
special circumstances, as when large storage of filtered water 
in service reservoirs is available, the clear-water basin may be 
made smaller. That at Albany, N. Y., holds only 600,000 
gallons, or less than one hour's supply. 
When a battery of several filters is under construction, it 
is very desirable that the separate beds be so arranged as to 
permit of the flow being watched from each one individually, 
otherwise the general filtrate might be damaged by the poor 
working of a single member of the group, and no means would 
exist of detecting and remedying the evil. 
In a paper before the American Water Works Association. 
March, 1914, Mr. G. A. Johnson gave sundry figures for 
comparative "first cost" and "cost of maintenance" of slow 
sand filters and "rapid" or "mechanical" filters (to be de- 
scribed later). 
"In discussing the cost of building water filtration works 
of the slow sand and rapid sand types, respectively, considera- 
tion will be given only to those items referring to the filter 
plant proper. Cost of land, pumping machinery, outside con- 
necting piping, intakes, etc., in fact everything outside the 
filtration plant proper, will not be considered. 
"For slow sand filter costs the items will include the nec- 
essary filter buildings and filters with all appurtenances, all 
inside piping, sand handling apparatus, preliminary sedimenta- 
tion basins, preliminary filters and appurtenances and clear- 
water reservoirs. 
"For rapid sand filter costs the items will include the filter 
buildings and filters with all appurtenances, all inside piping, 
filter washing apparatus, coagulating and clear-water basins. 
Thus a fairly good idea may be had of the relative cost of 
building purification plants of the two types. ARTIFICIAL PURIFICATION OF WATER 
133 
"It is true that, on account of the much greater area re- 
quired, the cost for land is far greater in the case of slow sand 
filtration systems than for rapid sand systems. Roughly, 
other things being equal, land will cost twenty times as much 
for a slow sand filter installation as for a rapid sand plant. 
Furthermore, in large projects, it is often difficult conveniently 
to locate a site for slow sand filters, while for a rapid sand 
filter plant it is a relatively easy matter as a rule. If it is 
necessary to go a long distance in locating an extensive and 
suitable area of land for a slow sand filter site there is incurred 
a large expense for a conduit to bring the filtered water to 
the city. This is very rarely necessary in the case of rapid 
sand filter projects. So that, in studying the comparative 
figures which follow, it must distinctly be borne in mind that 
the costs given for slow sand filter installations are really low, 
since the important considerations just mentioned are not 
charged against them. 
COST OF OPERATION AND MAINTENANCE OF SLOW SAND AND 
RAPID SAND FILTRATION PLANTS 
Year. 
City. 
Kind of Filters. 
Average Volume 
of Water 
Filtered Daily. 
U. S. Gallons. 
Cost of Operation 
and Maintenance 
per Million Gal- 
lons of Water 
Filtered. 
1911 
Albany, N. Y 
Slow sand 
20,000,000 
$2.50 
1912 
Pittsburgh, Pa 
Slow sand 
100,000,000 
3-41 
1911 
Philadelphia, Pa 
Slow sand (a) 
9,000,000 
5-62 
1911 
Philadelphia, Pa 
Slow sand (b) 
13,000,000 
3-59 
1911 
Philadelphia, Pa 
Slow sand (c) 
38,000,000 
3-88 ' 
1911 
Philadelphia, Pa 
Slow sand (d) 
202,000,000 
1.91 
1912 
Washington, D. C.. . . 
Slow sand 
62,000,000 
4.01 
1912 
Cincinnati, Ohio 
Rapid sand 
50,000,000 
4.12 
1911 
Harrisburg, Pa 
Rapid sand 
9,000,000 
3-93 
1912 
Little Falls, N.J 
Rapid sand 
30,000,000 
3.20 
1912 
Louisville, Ky 
Rapid sand 
25,000,000 
3-48 
1912 
New Orleans, La 
Rapid sand 
16,000,000 
6.32 
Weighted average | 
Slow sand 
Rapid sand 
$2.86 
4-04 
(a) Lower Roxborough; (&) Upper Roxborough; (c) Belmont; (d) Torresdale. 134 
WATER-SUPPLY 
"COST OF CONSTRUCTION OF SLOW SAND AND RAPID SAND 
WATER FILTRATION PLANTS 
City. 
Kind of Filters. 
Present Daily 
Filtering 
Capacity. 
U. S. Gallons. 
Approximate 
Cost per Million 
Gallons Daily 
Capacity. 
Albany, N. Y 
Slow sand 
20,000,000 
$20,000 (a) 
Pittsburgh, Pa 
Slow sand 
200,000,000 
26,000 (a) 
Philadelphia, Pa 
Torresdale 
Slow sand 
250,000,000 
37,7oo (a) 
Upper Roxborough 
Slow sand 
28,000,000 
29,800 
Lower Roxborough 
Slow sand 
17,000,000 
26,300 (a) 
Belmont . . 
Slow sand 
60,000,000 
45,200 (a) 
Washington, D. C 
Slow sand 
100,000,000 
30,000 (6) 
Cincinnati, Ohio 
Rapid sand 
112,000,000 
11,400 (c) 
Columbus, Ohio 
Rapid sand 
30,000,000 
13,000 (d) 
Dallas, Texas 
Rapid sand 
15,000,000 
13,000 
Harrisburg, Pa 
Rapid sand 
16,000,000 
10,300 
Little Falls, N. J 
Rapid sand 
32,000,000 
15,000 
Lorain, Ohio 
Rapid sand 
6,000,000 
14,000 
New Milford, N.J 
Rapid sand 
24,000,000 
11,000 
Watertown, N. Y 
Rapid sand 
8,000,000 
11,250 
Weighted averages ( 
Slow sand 
Rapid sand 
,$32,600 
12,100 
(a) Cost of preliminary filters included. 
(fe) Cost of Dalecarlia Reservoir not included. Cost of McMillan Park Reservoir 
included, and also cost of remodeling Georgetown Reservoir, as well as cost of coagulating 
basin. 
(c) Cost of large plain sedimentation basin not included. 
(d) Cost of softening works not included. 
"The above figures show that the approximate relative cost 
of building the slow sand and rapid sand filter plants mentioned 
was $32,600 and $12,100 respectively, per million gallons 
daily capacity. At 5 per cent the fixed charges on these sums 
would amount to $4.47 and $1.66, respectively, per million 
gallons of water filtered. 
"The cost of operation and maintenance of filtration plants 
in a large measure, varies, of course, with the quality of the 
raw water. In a general way the following examples will 
serve to show the charges ordinarily made against the opera- 
tion and maintenance of representative water filter plants 
in this country. 
"To summarize, the average cost of buildings even of the 
largest and most modern slow sand filter plants was $32,600 135 
ARTIFICIAL PURIFICATION OF WATER 
per million gallons daily capacity; and, likewise, the average 
cost of building six of the largest, and two medium size, rapid 
sand filtration plants was $12,100 per million gallons daily 
capacity. The average cost of operation and maintenance 
varied widely, of course, but averaged $2.86 and $4.04 per 
million gallons of water filtered by the slow sand and rapid 
sand systems, respectively. Adding these last figures to the 
fixed charge on the first cost of construction makes up the 
following totals: 
Slow sand filtration $7-33 per million gallons 
Rapid sand filtration 5-7° per million gallon s" 
The depths of water permitted upon filters of the slow 
sand type are almost as various as the compositions of the 
beds themselves. 
Three and a half to four feet of water may be taken as the 
depth most commonly in use, and it is important that this depth, 
when once determined upon, should be maintained a constant. 
In one of the older filters of this country the head of water 
varied exceedingly, running from a few inches to over six feet. 
Such inequality as must result in the rate of filtration cannot 
but cause wide differences in the purity of the filtrate. 
With the more modern filters the rate of delivery of filtered 
water is more independent of the depth of water on the sand, 
being controlled by an effluent regulator. A form devised by 
Lindley for the Warsaw works and shown herewith will serve to 
illustrate the principle. The outflow takes place through the 
horizontal slits under a head which is maintained constant by 
means of the floats attached to the upper end of the cylinder 
which telescopes the fixed outlet. 
By use of some such device as Lindley's, associated with 
a simple overflow-pipe to correct errors from too deep flooding 
of the bed, constancy in rate of filtration can be assured, a point 
of material importance in the proper management of a filter. 
The rate of filtration usually adopted in the past and still 
commonly in use is three million U. S. gallons per acre of 136 
WATER-SUPPLY 
sand surface per twenty-four hours. That is the figure generally 
held in mind when planning a slow sand filter, but the rate 
may be pushed to double that quantity if the character of 
EFFLUENT REGULATOR. (Lindley.) 
MIDNIGHT. NOON. MIDNIGHT. 
VARIATION IN HOURLY CONSUMPTION OF WATER AT BERLIN, STATED IN CUBIC 
metres. (Frankel.) 
the raw water, or its previous improvement by pre-filter devices, 
should warrant the higher rate. 
The tendency is certainly toward an increase in what, 
for convenience, may be termed the " standard rate," and ARTIFICIAL PURIFICATION OF WATER 
137 
it is therefore wise to make provision for a future greater 
volume of flow by designing the piping to carry the larger 
amount of water. 
The following are the rates at some American plants: 
Albany ' 3,000,000 U. S. gallons per acre per day 
Washington 2,500,000 " 
Pittsburgh 3,000,000 
Philadelphia (Belmont) 6,000,000 , 
Providence ' 2,000,000 " " " 
By direction of the Imperial Board of Health the maximum 
rate of filtration was fixed in Germany at four vertical inches 
per hour. 
The Massachusetts Board of Health finds that while such 
a rate is proper for new filters, a higher delivery may be as- 
signed to those which have been in service during a long 
period, " owing apparently to a more extended accumulation 
of gelatinous films within the main body of the filtering ma- 
terial." * 
When dealing with the question of the rate of filtration, the 
following table prepared by Mr. G. W. Fuller will be found 
convenient: 
EQUIVALENTS OF VARIOUS MEASURES OF RATE OF 
FILTRATION 
Million 
U. S. Gals, 
per Acre 
per 24 Hrs. 
U. S. Gals, 
per Sq. 
Foot 
per Hour. 
Cubic ft. 
per Sq. 
Yard 
per Hour. 
Vertical 
Velocity- 
in Inches 
per Hour. 
i million U. S. gals, per acre per 24 hrs. 
I. 
.96 
I-15 
i-53 
1 U. S. gal. per square foot per hour. , . 
I.045 
I. 
1.2 
1.6 
1 cubic foot per square yard per hour. . 
.869 
•83 
I 
i-33 
1 linear inch in vertical velocity per hour. 
•652 
.62 
•75 
1 
It will be observed that the rate of three million gallons 
per acre per day corresponds to a vertical velocity of 4.6 inches 
per hour. 
* Rep. Mass. Board Health, 1894, p. 609. 138 
WATER-SUPPLY 
CLEANING LONDON FILTER-BEDS. ARTIFICIAL PURIFICATION OF WATER 
139 
Cleaning a filter of the slow sand type is accomplished by 
a gang of laborers who work upon the drained bed, and by 
means of sharp shovels pare off the uper half-inch of sand and 
pile the same into small heaps, whence it is removed to the 
sand-washer. This thin upper layer of sand (the Schmutz- 
decke of the Germans) contains the greater bulk of the material 
separated from the water by filtration. It is quite compact, 
and in appearance is distinctly separated from the sand below. 
Its imperviousness to water is what causes the filter to become 
"dead" and require cleaning. 
The illustration given shows the cleaning of one of the 
London beds before the more modern methods were adopted. 
The Schmutzdecke (a portion of which is shown in place) was 
scraped into piles, as now, and such piles of dirty sand were 
then removed to the sand-washer by the use of wheel-barrows. 
After washing, the barrows were again employed to return 
the sand to the bed. 
The method of washing and handling sand is now one of 
greater expedition and less cost. 
Water under pressure is carried in a pipe suspended 
from the filter roof and is delivered by hose to a hopper 
containing an ejector equipment. Into this hopper the dirty 
sand is shoveled and is carried by the stream of water 
issuing from the ejector nozzle into another pipe, also sus- 
pended from the roof, which delivers it to some convenient 
point outside of the filter. The sand may be sufficiently 
scoured by such treatment and ready for re-use when the 
water, which carried it is drained away, or it may be re-washed 
if required. 
By the use of the "Nichols" washer the sand can be 
washed without leaving the filter. The "hopper" referred 
to above delivers into the Nichols apparatus, the cleaned 
sand is discharged directly from the machine and only 
the dirty wash water is carried off by the outflow pipe sus- 
pended from the ceiling. The washed sand is distributed and 
"surfaced" by hand. 
The "Blaisdell" sand washing machine used at Baltimore 140 
WATER-SUPPLY 
BLAISDELL EILTER-SAND WASHING MACHINE, BALTIMORE, MD 
(From Water Supply Paper. No. 315.) ARTIFICIAL PURIFICATION OF WATER 
141 
and elsewhere, is illustrated below and may be tersely described 
as possessing a washing chamber hanging from a traveling 
girder which spans the filter. This washing chamber may 
be raised and lowered so that it may be sunk unto the sand 
surface during use or be sufficiently elevated to clear the filter 
walls during transfer to another bed. Within the washing cham- 
ber is a stirring-wheel with hollow teeth, through which water 
under pressure enters the stirred-up sand. Suction is applied 
WASHING CHAMBER OF BLAISDELL FILTER-SAND WASHING MACHINE. 
" The chamber being slowly moved ahead, sliding on its shoes over the sand surface, 
water is forced through the hollow shaft to the hollow rim of the wheel, and thence to the 
hollow teeth below and through fine perforations into the sand. The dirty water from the 
washed sand is drawn away from the washing chamber through the suction pipe." 
The above illustration of a section is from a paper by W. B. Fuller in Engineering News 
March 12, 1908. 
to this chamber and the dirty water is thereby separated from 
the sand and pumped to the sewer. 
As a method for prolonging the life of a filter, Hardy reports 
excellent results were secured at Washington, by raking the 
sand surface of the drained filters with iron garden-rakes. 
Such raking "seemed to be nearly as effective in restoring the 
filter capacity as a scraping; it could be done in eight hours 
by three laborers, and there seemed to be no ill effects from 
lowered efficiency." The raking was first attempted from boats, 
first shutting off the effluent, but "this method was not satis- 142 
WATER-SUPPLY 
factory as the work was neither as uniform nor as thorough 
as necessary." * 
The "Brooklyn Method" of washing a slow sand bed con- 
sists in draining off the water to within a few inches of the 
sand surface, opening the outlets just above this surface, setting 
boards on edge so as to form a channel-way about 15 feet 
wide and then admitting a flow of wash water at one end of 
such channel-way with a velocity of half a foot per second. 
The upper sand layer is raked during the flow until the turbidity 
of the wash water is materially reduced. A lengthening of 
the life of the filter is thus secured. Because of the hard, 
pasty character of the material caught in the upper sand layer 
of the Torresdale filters, the "Brooklyn Method" was not there 
successful. 
The actual weight of dirt on dirty filter-sand is less than 
its looks would indicate. At Albany the writer found it to 
run about one-half of one per cent of the dirty sand dried at 
no0 C. 
Fine dirt will slowly find its way into the lower levels of the 
sand bed and consequently deep cleaning will be occasionally 
required. At Albany the body of the sand was cleaned for 
the first time after the filters had been running about ten years. 
Since then it has been cleaned about every four years. 
Excess of area, to allow normal filtration to continue while 
a portion of the plant is being cleaned, must be allowed for 
in the design. Middleton f considers that such area should 
run 16 to 25 per cent "of the area at work." This allowance 
is large even for an estimate made to suit a great variety of 
waters. About half of the higher figure would be more suit- 
able. The actual allowance made must of necessity depend 
upon the probable frequency of cleaning, which in turn will 
vary with the character of water to be filtered. 
It is better to design a filter of sufficient size, and construct 
it in a thoroughly up-to-date manner, than to expend the same 
sum of money for a larger but less efficient plant. 
* Am. Soc. C. E., 36: 1566. f " Water Supply," p. 115. ARTIFICIAL PURIFICATION OF WATER 
143 
The frequency of cleaning depends, as has been said, upon 
the condition of the water being filtered, and, for the same 
water, this condition will vary greatly with the time of year 
and character of the season. Thus the London beds experi- 
ence difficulty from March to July; becoming during those 
months quickly clogged with fish-spawn, which arrests filtration, 
and at times renders it necessary to remove the water from 
the top of the filter before the obstruction can be taken off 
with rake and shovel. 
From July until October another difficulty, scarcely less 
serious, is the growth of vegetation, which begins upon the 
bottom. 
Roughly stated, it may be said that a filter becomes "dead" 
(i.e., nearly impervious to water), and consequently demands 
cleaning, once every three or four weeks in summer and about 
half as frequently in winter; but it is not possible to lay down 
hard-and-fast general rules applicable to all filters. So many 
variable quantities enter the consideration that the ques- 
tion of proper time for scraping must be answered for each 
filter by itself, basing decision upon intelligent observation 
of the rate of flow, loss of head, and the obvious general 
efficiency. 
Where the ejector system is in use, it is not customary nor 
desirable to have the newly scraped filter remain out of service 
until the sand removed is washed and returned to its place; 
for the fine-sand layer should be thick enough to permit of a 
number of such successive scrapings without impairment of 
its efficiency; therefore the sand taken off during a series of 
parings is washed and returned at one time. 
At Albany the sand-layer, which is normally 48 inches thick, 
has been reduced as much as 17 inches by successive parings 
before the original thickness was restored by replacing the 
cleaned sand. 
The cost of cleaning a filter and washing and replacing the 
removed sand depends upon the plan adopted for the work, 
the design of the filter, and the local price of labor. 144 
WATER-SUPPLY 
The following figures give the cost for cleaning old-style 
open filters by the former "hand method": 
London (Southwark and Vauxhall).-To scrape and wheel 
dirty sand to washer requires 400 man-hours per acre of sand- 
surface. Labor, 9I cents per hour. Ice is not usually re- 
moved, but is raked into ridges. The cost of this is roughly 
found to be about three times that of the ordinary cleaning. 
London (New River).-To scrape one acre and wheel out 
dirty sand requires 144 man-hours. Labor, 9.3 cents per 
hour. 
Liverpool.-Cost of scraping and wheeling is $12.50 to 
$25 per acre, depending upon length of barrowing. Ice is 
occasionally broken up where it touches the walls, but is not 
removed. Labor, 9.1 cents per hour. Cost of cleaning per 
million U. S. gallons filtered, $1.14. 
Schiedam.-Scraping and wheeling out dirty sand requires 
174 man-hours per acre. Labor, 6 cents per hour. In Hol- 
land the sand is not usually washed, it being cheaper to obtain 
fresh sand for refilling. 
Rotterdam.-The mean cost for filter management, including 
cleaning, washing new sand, and trouble with ice, has been 
during ten years, $1.35 per million U. S. gallons water filtered. 
The following data for the Albany, N. Y., filter plant will 
serve for general illustration, although the plant is not the most 
modern one in the country. It was put in service in 1899 
and was then the last word in filter construction. Originally 
it consisted of only the covered slow sand beds, with a sedimen- 
tation basin attached. At present a pre-filter plant using alum 
is interposed between the sedimentation-basin and the filters 
proper, and a device for chlorination of the filtrate is erected 
at the entrance to the clear-water reservoir. 
Average dose of alum used 2 grains per gallon 
Average dose of available chlorine used. .0.49 parts per million 
The slow sand beds are eight in number, with a sand area 
of 0.7 acre each. The composition of the bed is: ARTIFICIAL PURIFICATION OF WATER 
145 
Gravel (egg size) 4 inches 
" (walnut) 4 " 
11 (marble) * 4 " 
(pea).... . ...4 " 
Sand (effective size 0.31 mm.) 40 inches (originally 48). 
The pre-filters are sixteen in number, with a sand area of 
820 square feet each. 
The sand of the pre-filters has an effective size of 0.65 mm. 
Capacity of clear-water basin, 600,000 U. S. gallons. 
Dimensions of clear-water basin, 94X94 feet. 
Elevation of sedimentation-basin above river, 18 feet. 
Dimensions of sedimentation-basin, 382.7X600X9 feet. 
Diameter of pure-water conduit to city pumping-station 
48-inch (steel riveted pipe). 
Depth of water on filter-bed, 4 feet. 
Total cost of construction of slow sand filter-beds, in- 
cluding piping and laboratory, $225,000. 
Cost of construction per acre net filtering area, $45,600. 
Cost of sedimentation-basin, $60,000. 
Cost of construction, pure-water reservoir, $9000. 
Cost of land, $8290. 
Cost of pure-water conduit and connections, $86,638. 
Cost of engineering and contingencies, $31,000 (original). 
Cost of pre-filters, $125,000. 
Total cost of entire plant, $700,000. 
Operation: 
Average total volume of water filtered per day, 21,444,000 
gallons. 
Average rate of filtration per acre per day, pre-filters. 
84,500,000 gallons. 
Average wash water used in pre-filters, 3.6 per cent. 
Average rate of filtration per acre per day, slow sand, 
2,998,000 gallons. 
Average depth of sand removed at each scraping, slow sand, 
0.62 inch. 
Average interval between scrapings, slow sand, 30 days. 146 
WATEK-SUPPLY 
Average interval between washings, pre-filters, 39.7 hours. 
Average gallons per acre between scrapings, slow sand, 
156,438,000. 
Permissible loss of head, 4 feet. 
Time required to scrape one filter, 30 man-hours. 
Time required to eject sand, 10 man-hours. 
Time required to level surface, 12 man-hours. 
Ten volumes of water are required to wash one volume of 
sand. 
Loss of sand by washing is five per cent. 
The table on page 147, from the Bureau of Water Report for 
1914, gives in detail the cost of purification by the Albany 
plant (quantity of water filtered, 7,817,285,000 gallons. Price 
of labor, $2.50 per day). 
At Torresdale, Pa., the most successful sand washing was 
done by the use of the Nichols machine. They found that the 
said machine used 1200 gallons of wash water per cubic yard 
of sand, as compared with 2800 gallons required by the ordinary 
ejector method. 
Hazen * considers that, as a general average, municipal 
filtration, excluding pumping but including all operating ex- 
penses and also including interest on cost of plant and allow- 
ance for repairs and depreciation, should cost about $10 per 
million gallons. 
A similar estimate was made by R. Winthrop Pratt of the 
Ohio State Board of Health, who says: f 
"It is estimated that filtered water, including interest 
charges and operating expenses, costs $6 to $10 per million 
gallons. This cost may be higher in special cases. At $10 
per million gallons, the cost to each inhabitant of a city where 
the daily water consumption is 100 gallons per capita, would 
be 0.1 cent per day, or 36.5 cents per year. This means an 
increase in the water rate of about ten per cent. Is not this 
an economical way of avoiding typhoid fever and similar 
* Trans. Am. Soc., C. E., 54; also see Whipple on " Value of Pure Water/' 
page 49. 
t Ohio Sanitary Bulletin, Jan., 1905. ARTIFICIAL PURIFICATION OF WATER 
147 
diseases? The price of fuel used in boiling water for drinking 
only would be much greater than 36.5 cents per year." 
COST OF FILTRATION AT ALBANY, N. Y. 
Pumping Station. 
Total 
Cost. 
Cost per Million Gals. 
$9,035 • 19 
791.95 
12,712.03 
546.70 
$1.16 
Incidental labor . . 
. 10 
Coal 
1.63 
Oil 
• 07 
3,438.86 
30 • 53 
• 44 
Ice. .. 
Total cost of pumping 
$22,555 ■ 26 
$3-40 
Filtration and Sterilization: 
Sedimentation Basin. 
Labor and teaming 
$226."5 
$226.75 
$0.03 
$0.03. 
Preliminary Fillers. 
Attendance and clerical work 
Removing, washing and replacing 
sand 
$3,230.36 
184.00 
280.60 
$0.41 
. 02 
Repairs, supplies, etc 
. 04 
$3,694.96 
$0.47 
Slow Sand Fillers. 
Scraping beds 
Rejecting scraped sand 
411.50 
617.75 
0.05 
. 08 
Washing and replacing 
2,198.50 
129.50 
. 28 
Removing ice 
. 02 
Reforking 
81.75 
.01 
Refilling beds 
830.50 
. 11 
Watchman 
729.00 
. 09 
Incidental labor 
4,493 • 58 
2,113 • 20 
69.75 
.58 
Repairs and supplies 
• 27 
Egg coal 
.01 
Ice 
30 • 53 
$11,705.56 
$1.50 
Laboratory. 
Chemist 
$2,100.00 
$0.27 
Laboratory help 
2,020.89 
. 26 
Supplies 
567.05 
• 07 
Gas-86° gasoline 
175.96 
. 02 
Egg coal 
69.75 
.01 
Ice 
30.54 
$4,964.19 
$2,595- 89 
$0.63 
$0.33 
Hypochlorale Plant. 
Attendance 
Repairs and supplies 
$908.60 
1,687.89 
$0.12 
. 21 
Alum Plant. 
Attendance 
$641.00 
8,032.85 
$0.08 
Repairs and supplies 
I • 03 
$8,673.85 
2,171.94 
$1.11 
.28 
Superintendence 
Total cost of filtration and steri'.i 
zation 
$34,032.34 
$4-35 
Total cost 
$60,588.40 
$7-75 
Cost per million gallons filtered 
$60,588.40 
7,817,285,000 
= 7-75. 
Before consenting to the large outlay of funds required for 
filtration by the English filter-bed system, the taxpayer very 
properly asks what efficiency may be looked for from such a 
construction. 148 
WATER-SUPPLY 
The best answer to this question is to quote the recorded 
duties of some existing plants. (Results in parts per million.) 
Color. 
Turbidity. 
Bacteria per cc. 
Albany, N. Y., 1914 (raw) 
33 . 
36 
66,850 
Albany, N. Y., 1914 (filtered).. . . 
22 
I 
29 
Providence, R. I., 1913 (raw).... 
46 
1,500 
Providence, R. I., 1913 (filtered). 
32 
IO 
Torresdale, Phila., 1914 (raw).. . . 
15 
20.2 
16,800 
Torresdale, Phila, 1914 (filtered) . 
IO 
0.6 
246 
Washington, 1915 (raw) 
0 
I90 
4750 (average) 
Washington, 1915 (filtered) 
0 
0 
12 (average) 
Pittsburgh, 1916 (raw) 
100 and over 
150 to 150,000 
Pittsburgh, 1916 (filtered) 
60% removal 
0 to 1 
Seldom over 10 
When the Albany filters were first placed in service their 
record was: 
Sept. 1899, Bacteria in raw water 16,800 per c.c. 
" " filtered water 343 per c.c. 
" removed 97-6% 
Sept. 1900. Bacteria in raw water 5000 per c.c. 
" filtered water 24 per c.c. 
removed 99.4% 
It is worthy of notice that the efficiency of the beds increased 
after the first few months of use. In other words, a new filter 
cannot be expected to render such good service as one which 
has become "ripened." 
The effect of water improvement by filtration upon the 
typhoid death-rate in New York State is well shown by the 
following data secured from the Health Department report. 
Average annual typhoid death-rate per 100,000 popula- 
tion in seven cities of New York before introduction of 
water-improvement (slow-sand filtration) 46.5 
Average typhoid death-rate for 1915 in same cities 13.3 
Similar data for eleven cities of the state before improve- 
ment by mechanical filtration 56.2 
Same for eleven cities for 1915 14.4 ARTIFICIAL PURIFICATION OF WATER 
149 
No better illustration can be given of what sand-filtration 
is capable of accomplishing than is presented by the record of 
the ten filters at Altona, Germany, during the month of Feb- 
ruary, 1893. The average number of bacteria per cubic centi- 
metre in the raw Elbe water for that month was 28,667, whilst 
the corresponding average for the filtered water was only 
90; showing a removal, by filtration, of 99.69 per cent of germs 
of all kinds. 
What this removal meant for the city of Altona during the 
cholera outbreak at Hamburg in 1892 has already been touched 
upon (page 49), and may be epitomized in a single word-■ 
11 safety." 
A similar illustration is to be found in the data from Cher- 
bourg, France. At the time of the epidemic of 1898-99 the 
city proper, of a population of 32,494, derived its supply from 
the river Divette, with imperfect purification by an old-type 
filter. At the same time the garrison numbering 8239, was 
supplied with water from the same source but without filtra- 
tion. 
The typhoid death-rates were: 
Civil 109 per 100,000 
Military 1893 per 100,000 
A second epidemic occurred in 1908-09. 
At that time the garrison supply was as before, but the 
city carefully filtered its supply by the Puech-Chabal system 
(the intake being as in the past). 
The typhoid death-rates were: 
Civil ' 6.5 per 100,000 
Military 1594.0 per 100,000 
There was one barrack within the city and no deaths oc- 
curred therein. 
Col. A. M. Miller, U.S.A., was the first to lay particular 
stress upon the necessity for a low number of bacteria per 
cubic centimetre yet remaining in the water after filtration, as 150 
WATER-SUPPLY 
together with a high percentage of removal, as properly rep- 
resenting a filter's efficiency. 
This seems a just view, for a high percentage of removal 
might yet permit of an entirely undesirable number of bacteria 
in the filtrate when operating upon a raw water of high pollu- 
tion. 
Given a raw water containing a large number of bacteria 
per cubic centimetre, it is a relatively easy matter to show a 
high percentage of efficiency in a filter used to purify the same; 
while on the other hand it is difficult to secure a high per- 
centage of germ removal, even with the best of management, 
if the "count" on the raw water be very low. 
A low residual number of bacteria should be required, 
rather than a high percentage of removal, and it is also right 
to demand the absence or practical absence of the Bacillus 
coli communis. 
If the Bacillus coli communis be absent in a filtered water, 
it is safe to assume the absence of the Bacillus typhosus also. 
The occasional presence of a few of the former is not greatly 
objectionable, for the reasons that it is much more numerous 
in even an infected raw water than is the Bacillus typhosus; 
it has a greater longevity under adverse conditions such as are 
furnished by water-carriage; and, further, it has been experi- 
mentally shown that, of the two forms, the Bacillus typhosus 
is the more readily removable by filtration; * not because of 
material difference in size, but rather on account of its rela- 
tively small power to resist the adverse conditions present 
in the filter. 
The thoroughness with which nitrification, and consequent 
purification, can be accomplished, by filtration through shallow 
beds of even very coarse material, has been worked out and 
demonstrated by the investigations of the Massachusetts 
experiment station, and one important point elucidated is that 
a small amount of oxygen (1 to 3 per cent) in the air of the 
filter is as effective as a larger quantity. 
The following extract from a report made some years ago 
* See Report Mass. State Board of Health, 1898, p. 487. 151 
ARTIFICIAL PURIFICATION OF WATER 
by Mr. G. W. Fuller, then in charge of certain experimental 
work at the station, will give an idea of the efficiency with 
which bacteria may be removed by five feet of filtering material 
intelligently managed: 
"The actual efficiency of the filters was tested by the 
application of typhoid-fever germs and other important kinds 
of bacteria, and observations on their passage through the 
filters. The pure cultures of the micro-organisms were grown 
in dilute bouillon solutions. Twenty-five or fifty cubic centi- 
meters of these solutious, containing millions of germs, were 
applied to the filters in a small quantity of water, and the 
effluent was examined at frequent intervals for several days. 
Fifty-five such experiments were made during the first five 
months of 1892, with these average results: 
Number of filters tested n 
Number of experiments with typhoid-fever germs.... 22 
Number-of experiments with B. prodigiosus 19 
Number of experiments with B. coli communis 14 
Average rate of filtration, gallons per acre, daily 1,350,000 
Number of bacterial determinations 914 
Average number of bacteria per cubic centimetre 
applied 104,200 
Per cent removed 99-48 
[The extreme limits in the rates of filtration in the several 
experiments were 280,000 and 2,600,000 gallons per acre, daily.] 
"These experiments were very severe tests upon the effi- 
ciency of the filters in removing bacteria, because the num- 
ber applied was probably greater than would occur in practice, 
and furthermore the organic matter introduced with the bacteria 
served them as a food material. The experiments made during 
the latter portion of the year are much fairer, because the 
bacteria were applied in small and long-continued doses at 
frequent intervals, and the food material applied with them did 
not increase the organic matter in the river-water beyond the 
limits of variation observed from time to time in the amount 
originally present. The species of bacteria used was B. pro- 152 
WATER-SUPPLY 
digiosus, on account of its easy and reliable differentiation, its 
similarity to the typhoid-fever germ in its mode of life in 
Merrimac River water, and the fact that it has never been 
found native in this country. In the following table are sum- 
marized the average results of daily experiments from September 
16 to December 31, 1892: 
Number of filters experimented upon 11 
Number of bacterial determinations 2,372 
Rate of filtration, gallons per acre, daily 1,700,000 
Average number of B. prodigiosus per cubic centi- 
metre applied 5,700 
Per cent removed. 99.87 
The necessity for filtering to waste for a time after cleaning 
a slow-sand filter, as indicated by the earlier Altona results, 
does not accord with the experimental experience obtained 
at the Lawrence experiment station,* nor can it be classed as 
sound modern practice. If the scraped bed be carefully filled 
to above the sand-surface with clean water admitted through 
the underdrains, and fdtration be then resumed at a moderate 
rate, which is gradually increased to the normal flow, there 
is no reason for obtaining any other than good results. The 
necessitate for wasting the filtrate after ordinary scraping 
has not been felt at Albany. 
It is proper to require a slow and careful filling of the sand- 
voids with the water admitted through the underdrains, for if 
such water be passed in hurriedly, disturbance of the sand body 
by the escaping air will take place, resulting in the passing 
of bacteria through the channelways so formed. 
Especial care should be taken that no freezing of the sand 
take place during the time of cleaning. The difficulty of pre- 
venting this is, of course, a question involving climate, and 
for very cold countries the only solution of the problem is the 
construction of covered filters. 
* Report Mass. State Board of Health, 1893. ARTIFICIAL PURIFICATION OF WATER 
153 
Where open filters are in use in Europe, the attendants 
depend upon a careful watching of the weather, and very rapid 
work during cleaning, to prevent freezing. Should the sand 
once become frozen, the difficulty is a serious one, for, as was 
demonstrated at Altona, the water above the sand does not 
thaw the ice nearly as rapidly as one would expect, and more- 
over, thawing takes place unequally over the surface, whereby 
a discharge of the whole filtrate occurs through only a frac- 
tion of the bed, with most unsatisfactory results. 
A recent critic of the filter-bed system refers to the removal, 
by the bed, of ninety-odd per cent of all bacteria present in 
SKETCH SUGGESTING RELATIVE SIZES OF BACTERIA AND SPACES BETWEEN 
SAND GRAINS. 
the raw London water, and then notes that reduction of' the 
typhoid death-rate has been by no means in so large a pro- 
portion. As an explanation of the observed facts, he suggests 
that the small size of the typhoid bacillus prevents its being 
arrested by the filter as* easily as its larger companions. 
In reply it must be said, firstly: Typhoid fever would not 
entirely disappear from London, or any other city, were the 
water-supply absolutely sterile, for the sufficient reason that 
bad water, although a main cause of the disease, is not the 
only one. Secondly: Filtration is very far from being a simple 
straining process; were it so, all bacteria would pass the filter 
with equal facility, for differences between the several sizes 
of these extremely minute objects would be small indeed as 154 
WATER-SUPPLY 
compared with the spaces between the grains of sand. What- 
ever its size, the bacterium is caught by the zoogloea jelly sur- 
rounding the grains of sand, or is killed by the adverse con- 
ditions set up during nitrification, or both. 
It is worthy of note that the Massachusetts Board of 
Health * found that if highly polluted water be filtered and the 
filtrate be then passed through a second filter, the efficiency of 
such second filter will be low and remain low for a long time, 
owing to the lack of the gelatinous coating on the sand grains. 
The character of sand most suitable for water-filtration 
now receives decidedly more attention than it did in the past, 
for it is recognized that not only must a proper size of grain be 
secured, but that the degree of uniformity in size must be 
considered as well An ideal sand would be one having all 
of its grains exactly of one size, therefore it is sought to have 
the "uniformity coefficient" as near unity as possible. 
Moreover, it is to be remembered that much carbonate 
of lime present in the sand will harden the water filtered 
through it. 
It is easy to determine the presence of carbonate of lime, 
by observing whether or not the addition of hydrochloric acid 
to the sand causes an effervescence, but an accurate knowledge 
of the amount of the carbonate present calls for a chemical 
analysis. 
A large amount of work has been done at the Lawrence 
experiment station (notably by Hazen) upon the efficiency of 
sand,f and the expressions there employed are now widely 
adopted. 
The "effective size" of a sand is that diameter of grain 
than which ten per cent of the sand, by weight, is either equal 
to or less. It is expressed in millimetres. 
If there should be also determined that diameter of grain 
than which sixty per cent of the sand, by weight, is either 
equal to or less, and if this value be divided by the "effective 
* Report of 1896, p. 524. 
t See Reports of Mass. State Board of Health, 1892, p. 541, and 1894, p. 703. ARTIFICIAL PURIFICATION OF WATER 
155 
size" already found, the quotient will be the "uniformity 
coefficient." 
The mechanical analysis of sands is most easily accom- 
plished by the use of standardized sieves. Those used by 
the writer are of brass of the following sizes and are "nested": 
No. of Mesh. 
True Size in 
Millimetres. 
IOO 
O.l6 
•So 
O.I9 
60 
O.27 
40 
O.46 
20 
0.88 
l6 
1.16 
IO 
2.04 
8 
2.74 
It is possible to purchase sieves of the larger sizes accurately 
made by boring plate metal with standard drills; but the smaller 
sizes must be woven, and therefore have to be standardized. 
This is best done by the method described in the Massachusetts 
State Board of Health report for 1892. 
For each sieve a definite number of those sand grains are 
taken which are the last to pass, and which consequently most 
nearly correspond in diameter with the mesh of the sieve. 
After counting, these grains are weighed in milligrams, and 
their total weight divided by their number will give the weight 
of one. The specific gravity is easily taken by the bottle 
method. 
Now, since the volume of a sphere is equal to ^ird3, and 
also since W - V X specific gravity, we have 
x ,3 * 
in which d, the diameter of a sphere equal in bulk to the sand 
grain, is the only unknown quantity. 
Thus the true size of each mesh is secured and recorded. 156 
WATER-SUPPLY 
The mechanical analysis of sand is accomplished as fol- 
lows: Take 100 grammes of the sample. Place this in the 
top of the "nest" of sieves, thoroughly shake and weigh the 
PER CENTS CY WEIGHT 
DIAMETERS IN MILLIMETERS 
DIAGRAM OF MECHANICAL ANALYSIS OF SAND. 
material passing each sieve. Convert these weights into per- 
centages of the original weight taken and plot them as illus- 
trated above. The data for the curve there given are as 
follows: ARTIFICIAL PURIFICATION OF WATER 
157 
Size of Sieve in 
Per Cent of 
Millimetres. 
Sand Passing. 
o. 16 
0.0 
0.19 
0.0 
0.27 
0.1 
0.46 
• • 5 
0.88 
96.0 
1.16 . . . ? 
99-4 
2.04 
99 •6 
2-74 
100.0 
It will be observed that the curve cuts the ten per cent 
line at .45 and the sixty per cent line at .76. Hence the 
"effective size" of the sand under examination is .45 and the 
"uniformity coefficient" .764-45 = 1.6. 
Data as determined by Mr. H. W. Clark for sand from 
some foreign filters are here given: 
Effective 
size. 
Uniformity 
coefficient. 
Hamburg 
o-3i 
2-3 
Altona 
°-37 
2.0 
Berlin 
o-35 
i-7 
Zurich . 
0.28 
3-2 
Liverpool 
°-32 
2-5 
Other data are: 
Washington 
o-32 
1.64 
Albany 
0.31 
2-3 
Pittsburg 
0.29 
2 . I 
Philadelphia 
28 to .36 
2-5 
A good general specification for sand would be: 
Effective size, 0.35 m.m. (about); 
Uniformity coefficient, 1.5 m.m. (about). 
In an article on the New Haven plant Mr. C. A. Ferry says: * 
"Our specifications for sand were: Material which will 
pass a screen with 0.1-in. mesh and will not contain more than 
* Eng. News, May 9, 1907. 158 
WATER-SUPPLY 
0-5 per cent of material which will pass a screen with o.oi-in. 
mesh. It shall have an effective size of not less than 0.25 m.m. 
nor more than 0.35 m.m.; and shall have a coefficient of uni- 
formity of nor more than 2|." 
"rapid" or "mechanical" filtration 
Roughly outlined, this system consists in adding to the 
water to be filtered a minute dose of some "coagulating" 
material, usually common alum, or better, aluminum sulphate, 
averaging about one grain per gallon, and then after a short 
sedimentation period, admitting the water to the filter, which 
is a tank of wood, iron, or concrete, containing a filtering bed 
consisting of about 27 inches of sand (.40 to .65 m.m. effective 
size) resting upon about 9 inches of gravel. Above the sand 
surface the water stands about 24 inches deep. The alkalinity 
naturally present in the water decomposes the alum, with the 
formation of a white flocculent precipitate of aluminum hy- 
droxide, quite jelly-like in appearance. For instance, the car- 
bonate of calcium, representing the alkalinity, acts as follows: 
Al2(SO4)3+3CaCO3+3H2O = 2Al(OH)3+3CaSO4+3CO2. 
The action of this aluminum hydroxide is much the same as 
that of the white of egg in clearing coffee. It entangles all 
suspended matter, bacteria as well as inorganic material, and 
deposits the same on the surface of the sand, whence it is 
removed and driven into the waste-pipe by a reverse current 
of filtered water at the time of cleaning the filter. Thus, it 
will be observed, the mechanical filter produces an artificial 
inorganic jelly to replace the "bacterial jelly" of the slow 
sand filter-bed, already alluded to on page 119. 
A further action of the precipitated aluminum hydroxide 
is to unite with the soluble coloring matter of the water, whereby 
the filtrate is rendered colorless. The proper "dose" of alum 
solution is administered by means of some small automatic 
measuring apparatus exterior to the filter; the liquid flowing ARTIFICIAL PURIFICATION OF WATER 
159 
thereto from the dissolving tank, where the water solution of 
the coagulant is prepared of convenient strength. 
AUTOMATIC PRESSURE-FILTER. 
For small establishments it is a convenience to make use 
of a 11 closed" or "pressure" filter similar to that illustrated 
on this page, for the reason that it can be placed directly upon 160 
WATER-SUPPLY 
the service-main, thus allowing no interruption in the line 
of the supply. It is not easy, however, to carefully watch such 
a filter, and it is open to sundry other objections, notably a 
lack of uniformity in action. For purposes of public service 
there is some prejudice to its use as a substitute for the open or 
11 gravity" type. 
On pages 161 and 170 are given illustrations of such "open" 
or "gravity" mechanical filters of the tub form, still in use for 
municipal service, but commonly replaced in larger plants by 
rectangular concrete tanks such as shown on page 168. "Units" 
are erected in numbers and sizes suited to the quantity of water 
to be filtered. 
The rale at which water is passed through these filters 
is very great when compared with what is furnished by the 
English filter-beds already considered. One hundred and 
twenty-five million gallons per acre per twenty-four hours 
may be taken as a fair index of what is to be expected of 
them. 
It has been already said that alum is used in their opera- 
tion. The necessity for the use of this or some other coag- 
ulant is imperative and should be impressed upon the mind, 
for, however well the straining action of the sand-bed may 
clarify the water, the great rapidity of flow precludes the 
proper removal of bacteria without they be entangled in the 
alumina jelly noticed above. 
Alum * is the principal, but by no means the only "coag- 
ulant" that is used in mechanical filters. .Salts of iron were 
suggested some years ago as a cheap substitute for alum and 
are so employed, especially in the Middle West, but their use 
is not greatly extended. 
Large plants using ferrous sulphate and lime are in suc- 
cessful operation, notably at Cincinnati and New Orleans. 
The ferrous sulphate is employed to replace the alum, but 
as it is not quickly acted upon by the natural alkalinity of 
the water (as is the case with alum), the addition of lime 
* Used henceforth to denote basic aluminum sulphate. ARTIFICIAL PURIFICATION OF WATER 
161 
is required to cause the rapid formation of the hydroxide of 
iron. 
At Loraine, Ohio, scrap iron was used. Sulphurous acid, 
formed by burning sulphur and admitting the fumes mixed 
with steam to a condensing-chamber, was allowed to act upon 
the iron scrap in a revolving drum. The iron sulphite formed 
"ofen" or "gravity" mechanical filter, showing sedimentation- 
chamber BELOW SAND-BED. 
was subsequently decomposed with lime, and the resulting 
iron hydroxide acted as a coagulant in a manner entirely similar 
to the corresponding compound of aluminum. 
The advantage claimed for the iron process is economy 
in cost of coagulant, and rapid settlement of the flock, but the 
expense of application is increased, which partly offsets the 
lower first cost, and moreover, the method does not work well 
in some instances. Iron salts will react with the organic color- 162 
WATER-SUPPLY 
ing matters in certain peaty waters with the formation of soluble 
compounds that increase the color of the filtrate. 
As already stated, the municipal plant of Cincinnati, Ohio, 
filters the Ohio River water by use of the iron and lime process 
and the results are certainly successful. The following data 
for that plant are for 1913: 
Capacity, 112 million gallons daily. 
Dose of iron sulphate 1.6 grains per gallon (costing $12 per 
ton). 
Dose of lime, 0.81 grains per gallon (costing $5.69 per ton). 
Dose of chlorine (liquid) .12 parts per million. 
Filters have 30 inches of sand, 0.32 m.m. effective size, 
resting on V-shaped under-drains 8 inches deep and filled with 
gravel | to 1 inch size. 
Raw water is pumped to settling-basins, whence it flows 
to the filters, developing electric power by its fall. 
Time of settling in such basins, about forty-eight hours. 
Coagulation period averages nine hours. It is never less 
than 4I hours. 
Average time between washings, 17I hours. 
Average time to wash a filter, 4.15 minutes. 
Average wash-water used, 2.65%. 
Average bacteria per c.c. in raw water, 16,100. 
Average bacteria per c.c. in filtered water, 56. 
Average bacterial efficiency, 99.7%. 
Cost per million gallons water delivered, $3.92. 
Of this $1.67 was for chemicals. 
Based on the figures given by J. W. Ellms, the above 
figure for cost of chemicals would increase to about $3.05 if 
alum were used. 
At New Orleans in 1910 the record stood: 
Lime, 4.8 grains per gallon. 
Ferrous sulphate, 0.76 grain per gallon. 
Average period between washings, 136.6 hours. 
Wash-water required, 0.75 per cent. 
Rate of filtration, 59.9 million gallons per acre per day. 163 
ARTIFICIAL PURIFICATION OF WATER 
Cost per million gallons: 
Chemicals, supervision, and labor $6.60 
Wash-water and reservoir cleaning 0.17 
$6.77 
The plant is of peculiar type and not strictly comparable 
with mechanical plants generally. * 
The plant at Cohoes, N. Y., is a small one of only eight 
million gallons capacity, but it is modern and will serve as a 
good illustration of an alum-using filter. The water is pumped 
from the Mohawk River to a sedimentation reservoir, whence 
it flows to the plant, receives its dose of alum on entering, 
passes to the 11 coagulation" or "reaction" basin and then 
on to the filter house. Chlorination is accomplished by the 
use of "bleaching powder" (so-called "hypochlorite of lime") 
which is added to the water after it has passed the filters. 
The filters are eight in number, each of one million gallons 
capacity. The filter-beds consist of 27 inches sand, resting 
on 4 inches |-inch gravel, three inches |-inch gravel, and two 
inches |-inch gravel. The water above the sand is 24 inches 
deep. 
The dose of alum is about one grain per gallon usually, 
although at times rises to twice that quantity. 
The dose of bleach is 0.3 part per million of available 
chlorine. 
Cost of alum, $1.05 per hundred pounds. 
Cost of bleach, $1.95 per hundred pounds. 
Time between washings, twelve hours. 
Coagulation period, two to four hours. 
Bacteria in raw water, 20,000, 1912. 
Bacteria in filtered water, 5, 1912. 
What amount of coagulant should be used is a question 
best answered by the attendant in charge; for not only will 
* See Engr. Record, 64: 356. 164 
WATER-SUPPLY 
the proper "dose" vary with different plants, but even in 
the same plant it may differ from day to day, and certainly 
will change from season to season. One grain per gallon is 
a fair estimate for a rough one. This amount may be often 
somewhat lessened with satisfactory result, but also must 
frequently be materially increased. 
It should be remembered that good results are what are 
looked for, and that if they be not obtained by the dose of 
coagulant adopted, such dose must be gradually increased until 
the maximum amount allowable shall have been reached. 
It is true, as may be seen from the equation already given, 
that the use of alum increases the permanent hardness of the 
water because of the formation of calcium sulphate; but the 
extent of the hardening is commonly insignificant, as an ob- 
jection, compared with the outlay caused by heavy doses of 
the coagulant. 
The influence of turbidity upon the quantity of alum required 
is an interesting matter worked out by Fuller in connection 
with the Louisville experiments. It seems that marked tur- 
bidity calls for a dose of coagulant entirely beyond what the 
soluble salts present in the water demand; and that if the 
additional dose fall short of what is required by the " absorptive 
power" of the suspended matter, coagulation fails and the 
chemical is practically wasted. In view of this, Fuller dwells 
upon the economic value of sedimentation as a preliminary 
to filtration of very turbid water; * the good point is further 
made that when the water (Ohio River) is muddy, and con- 
sequently requires an extra amount of coagulant, with its 
attending increase of boiler-incrusting salts, the raw water is 
then so low in such substances as produce hard scale that the 
total incrusting power of the filtrate is less than that of the 
raw water when at its clear stage.f 
It is worth remembering that theoretically a "dose" of 
one grain per gallon of aluminum sulphate (containing 17 per 
* Louisville Report, p. 386. 
f Ibid., 248. ARTIFICIAL PURIFICATION OF WATER 
165 
cent AI2O3) will require about 7I parts per million of " alka- 
linity" in the water for its complete decomposition. 
The chemicals required for mechanical filtration are usually 
prepared in strong solution, or suspension, in suitable tanks 
and are supplied thereform in liquid form to the raw water. 
"Dry-feed" is, however, employed occasionally, the chemical 
in solid form and known weight being added to a tank through 
which water is flowing in an upward direction on its way to 
the main raw supply. 
In plants using lime, or "bleach," large doses of the chem- 
ical may readily form hard deposits in the bed, especially in 
cold weather when carbon dioxide is high in the water. A 
mass of " conglomerate," due to this cause, and consisting of 
underdrain gravel cemented together by calcium carbonate 
formed in the bottom of one of the Cohoes filters during the 
winter of 1915. 
It is common to write specifications demanding that " Filter 
Alum" shall contain not less than 17.5 per cent of AI2O3. 
Whipple and Longley * have shown that a basic alum 
CONSOLIDATION OP FILTER-BED, COHOES. 
* J. Infectious Diseases, Sup. 2, Feb., 1906. 166 
WATER-SUPPLY 
(i.e., one containing an excess of AhOs, as is commonly the case 
if that item be over 17.5 per cent) will produce a more rapidly 
forming and better flock than will be given by a normal alum. 
C. P. Hoover, Chemist to the Columbus, Ohio, Filter Plant, 
in October, 1915, patented a process for supplying "filter 
alum" at greatly reduced cost by simply boiling bauxite in 
sulphuric acid, in proportions suitable for the formation of a 
basic aluminum sulphate; adding water until the dilution is 
equivalent to the alum "feed" commonly used and then, 
without filtration, employing the solution with its contained 
"chemical mud" for addition to the raw water in the usual 
way. Not only is there a saving of "five distinct steps in 
alum making, namely, filtering, concentrating, crystallizing, 
grinding, and redissolving," but the "mud" itself is claimed 
to possess the advantage of furnishing nuclei for starting coag- 
ulation. 
Considerable discussion has arisen concerning the arsenic 
content of filter alum,* the arsenic being derived from the arsen- 
ical pyrites from which the sulphuric acid used in the man- 
ufacture of the alum was made. The amount of the arsenic 
present is, however, altogether too minute to be of moment. 
Thus, as Bartow and Bennett have shown, assuming the water 
to have received the large dose of six grains of alum per gallon, 
and assuming that all of the arsenic remained in solution in 
the filtrate, which is a practical impossibility, then " since 
arsenic is not a cumulative poison, a person must drink 1285 
gallons of the filtered water at one time to obtain a minimum 
medicinal dose of arsenic." 
Danger to health from the use of alum is a topic fruitful of 
discussion among the many who are not posted as to its manner 
of action; but those who are better informed know that free 
alum never reaches the filtrate in a well-ordered plant. 
Alum in the filtrate means a useless waste of material, not 
to be excused. Reference to the equation already given shows 
that in order to produce the aluminum hydroxide jelly, upon 
* J. Am. Water Works Asso., 2: 585. ARTIFICIAL PURIFICATION OF WATER 
167 
which successful filtration depends, the alum admitted to the 
water must be entirely decomposed, and never reaches the 
water consumer. 
MECHANICAL FILTER PLANT, NORFOLK, VA 
The method of testing for free alum is very simple, and the 
filter attendant, constantly on watch for it in the filtrate, should 
so manage the "dose" employed as never to allow it to pass 
beyond the sand-bed. 168 
WATER-SUPPLY 
The illustration on page 167 gives an idea of how a collec- 
tion of the filtering units of the tub form is set up. 
That given below shows the inside of the filter-house connected 
MECHANICAL FILTER PLANT, ERIE, PA. 
with a group of rectangular concrete filter-tanks at Erie, Pa., 
a form commonly built for large works. It will be observed 
that only the ends of the filter tanks extend into the house, 
the remainder of them being covered by the outside terrace. ARTIFICIAL PURIFICATION OF WATER 
169 
Washing filters of the type illustrated on pages 161 and 167 
is assisted by revolving rakes, driven by power. A reverse cur- 
rent of clean water, entering through the bottom valves, loosens 
up the entire depth of sand and causes the wash-water, with its 
contained dirt, to overflow the rim of the drainage gutter and 
thence passes to the sewer. 
The amount of wash-water used for purposes of such cleaning 
is an item for consideration, for it is, of course, lost to the regular 
volume of filtered supply. 
The quantities used at sundry plants must naturally vary 
with local conditions, but a fair general statement will place 
it at from 2 to 3 per cent of the total water filtered. 
In plants of the types pictured on pages 168, 170 and 171, 
stirring rakes are omitted and the required scouring of the 
sand is secured by driving air up through the sand-bed at 
a pressure of about five pounds per square inch. This "air 
wash" is-continued for about three minutes, after which the 
wash-water is driven up through the under-drains as usual, 
for a period of about four minutes, and the dirty water escapes 
through gutters leading to the sewer. The entire length of 
time during which the filter is out of commission due to the 
operation of washing is thus only a question of a few minutes. 
It is not desirable to make a thorough removal of all dirt 
by the washing process for the reason that a portion of the 
"catch" of the previous run forms a sort of thin Schmutzdecke 
as it falls back upon the sand-surface at the end of the wash 
and thereby increases the efficiency of the filter. 
It is more common than not to have the compressed air 
used for the "air-wash" driven through the same valves in 
the under-drain system that are used to carry off the filtered 
water; but in some designs there are special valves used for 
the passage of air alone. This arrangement is somewhat more 
expensive as to first cost. 
Mr. J. W. Ellms has shown that the "air-wash" and me- 
chanical stirrers may be dispensed with in cleaning, provided 
that the upward flow of wash-water can be increased in velocity 170 
WATER-SUPPLY 
to eighteen inches per minute as measured above the sand- 
surface. Mr. W. Wheeler patented a form of bottom and under- 
CONTINENTAL AIR WASH GRAVITY FILTER. 
drains designed to allow of increased velocity of wash-water and 
of its uniform distribution.* This is illustrated on page 172. 
After receiving its dose of alum, but before the filter-sand 
* See also paper by R. S. Weston, Eng. News, July 2, 1914. ARTIFICIAL PURIFICATION OF WATER 
171 
is reached, the water passes through a "coagulation" or "reac- 
tion" chamber, depending in size and shape upon the design 
PORTSMOUTH, OHIO, FILTER PLANT. 
(Showing " air-wash " in operation) 
of the plant, wherein opportunity is given for sedimentation 
of a portion of the heavy magma produced by coagulation; 
otherwise the sand-bed quickly chokes and the labor and 
expense of cleaning are greatly increased. 172 
WATER-SUPPLY 
//i .T. 
3x|4 extending beneath 
i 2 Floor Plank 
^Flat Top & Bottom Tie Rods \t. 
Washer Intermediate Tie Rods 
Sides and Bottom 
of 2'Surfaced Plank 
VERTICAL SECTION A-B 
THE WHEELER FILTER BOTTOM. 173 
ARTIFICIAL PURIFICATION OF WATER 
As is seen from the sectional cut on page 161 a chamber is 
provided in some filters of the mechanical type whereby sedi- 
mentation advantages are to an extent secured even when 
the water is in rapid motion. 
For the information of those unacquainted with this type 
of mechanical filter, let it be said that each "unit" (of which 
there are eighteen in the battery at Elmira) consists of a tub- 
shaped vessel about 14 feet high, and open at the top. 
The bed of filtering sand, 3 feet 6 inches deep, together 
with the washing machinery, occupy the upper portion of the 
tank, leaving below an empty chamber 6 feet high, through 
which the raw water must pass on its way towards the sand. 
A test was made to determine what this lower chamber 
would do in the way of relieving the filter proper of a portion 
of its work. The water, carrying its dose of alum, entered 
in a tangential direction, near the circumference of the circular 
bottom, and flowed to the sand-bed through a pipe in the centre 
of the roof. 
Such an arrangement naturally caused the water to take 
up a movement of rotation, more or less coincident with the 
surface of a cone whose base was the floor of the chamber 
and whose apex was the orifice in the roof. 
Upon opening the chamber at the end of a seven days' 
run, a large quantity of aluminum hydroxide was found lying 
upon the floor in forms like drifted snow and varying in depth 
from almost nothing near the circumference to 14 inches and 
more towards the centre of the bottom. Its volume was equiv- 
alent to many cubic feet. The best idea of the amount of 
deposit and of its appearance is given by the accompanying 
flashlight photograph (see page 174), which also shows the 
uprights, graduated to inches, standing therein. 
The photograph shows but poorly one other feature of 
this aluminum hydroxide deposit, which was of considerable 
interest, namely, the numerous ball-like aggregations occurring 
throughout the mass. It would seem that the hydroxide had 
in many places been rolled together and built up much after 
the same manner that a boy accomplishes the gradual enlarge- 174 
WATER-SUPPLY 
ment of a ball of snow. This formation resulted from the 
rotary motion given the water by the special device for its 
introduction already alluded to. Such a motion is favorable 
DEPOSIT OF ' COAGULATED " MATERIAL IN SETTLING-TANK OF MECHANICAL FILTER. ELMIRA PLANT. 
to the formation of large flocks in a liquid in which aluminum 
hydroxide is already formed. Certainly it is a, mistake to 
suppose, as has been so often maintained, that the most rapid 
"coagulation" and precipitation of the hydroxide is neces- ARTIFICIAL PURIFICATION OF WATER 
175 
sarily to be secured by permitting the liquid containing it to 
remain in a state of complete rest. The photographs (on page 
176) illustrate that gentle rotary motion, far from being objec- 
tionable, is a distinct advantage. The two jars there shown 
differ from each other but in the fact that the left-hand one 
had its contents gently rotated, while the contents of the com- 
panion was maintained at complete rest In each the same 
quantity of hydroxide had previously been formed. The photo- 
graph indicates how progressively advantageous was the rotary 
motion for the formation and deposit of large flocks of the 
hydroxide; nor is this to be wondered at when one considers 
that the movement in question furnished opportunity for more 
frequent collision, with resulting cohesion, of the growing 
flocks. 
With reference to the bacterial efficiency of these tanks, it 
was interesting to observe that at Elmira an average of 42.6 
per cent of the bacteria present in the raw water had been re- 
moved by the time the sand was reached. 
As a further inquiry a closed pressure filter, operating 
without a chamber between it and the point of introduction 
of the alum, was tapped, both above and below the sand-sur- 
face, whereby it was possible to secure samples of water from 
various depths in the sand body. 
Results showed that the alum was not entirely decom- 
posed, and consequently the chemical equation not fully sat- 
isfied until considerable depths in the sand-bed had been pene- 
trated. 
This was to have been expected, in view of the fact that 
time is an element in all chemical reactions, and a pressure 
filter, running under the conditions named, allows but small 
time to elapse after the addition of the alum before the sand 
is reached. 
As already said, the precipitated aluminum hydroxide has 
a marked tendency to unite with soluble coloring matter, which 
fact makes the alum process of distinct value for removal of 
the yellow stain so often seen in surface-waters. 176 
WATER-SUPPLY 
PROGRESS OF SUBSIDENCE TN ROTATED AND IN QUIET WATER. ARTIFICIAL PURIFICATION OF WATER 
177 
When the water is so soft as to be deficient in the amount 
of alkalinity necessary to decompose the alum added, either 
lime or "soda-ash" is added to the raw water. 
These two chemicals must not be considered as always 
identical in action. "Soda-ash" (which is a carbonate) is 
most commonly depended upon to furnish the lacking alka- 
linity, but sometimes it is disappointing; the case apparently 
demanding a reagent containing the hydroxyl ion which is 
furnished by slacked lime. 
It may well be, therefore, that the failure following the 
use of one of the above chemicals will turn to success upon 
substituting the other. 
When doing experimental work to determine the best 
method of treatment for some particular water, it must not 
be forgotten that, if small experimental filters are used, the 
annular space between the filter wall and the sand-bed 
bears a much larger relation to the filter area than is the 
case with the full-size plant. Consequently a larger per- 
centage of imperfectly treated water must pass into the 
filtrate. 
The cost of erecting a mechanical filter-plant is somewhat 
difficult to give, for the reason that patent rights enter the 
estimate. It is true that the Hyatt patent for the use of alum 
has expired, but there are sundry devices dealing with im- 
provements in the filtering apparatus itself which are pro- 
tected by patents, and which it might be inconvenient to do 
without. 
Perhaps the fairest general statement that can be made 
is to place twelve thousand dollars per million gallons daily 
capacity as the basis price for a first-class plant, and to alter 
that figure as local conditions may demand. Reference to 
costs will be found on page 134. 
Filters can be erected for half the above amount, and they 
do good work for manufacturing purposes, but their bacterial 
efficiency is low and the water used for washing is higher than 
in the best forms of apparatus. 178 
WATER-SUPPLY 
The cost of operating a mechanical filter-plant must of neces- 
sity be a variable quantity, but it will run not very far from 
four dollars per million gallons of water filtered. The value 
given in Johnson's estimate will be found on page 133. A 
recent figure for the large Cincinnati plant is $3.92. These 
figures do not include interest on first cost. 
The efficiency of mechanical filters in the matter of bacteria 
removal is practically the same as for the slow-sand type, 
assuming each to be carefully run. Mr. S. T. Powell, in charge 
of the Avalon plants, of the Baltimore County Water Co., 
operates both types of filters with the same raw water. 
He reports the following averages for eleven months of 1911: 
Type of Filters. 
Average Bacteria 
per Cubic Centi- 
metre, Unfiltered 
Water. 
Average Bacteria 
per Cubic Centi- 
metre Filtered 
Water. 
Average per Cent 
of Bacteria Re- 
moved by Filters. 
Mechanical -. 
2414 
8 
99-57 
Slow-sand 
24M 
II 
99-55 
With regard to turbidity removal, and especially the elim- 
ination of color, the results secured by the mechanical plants 
are superior. 
"Pin-point" coagulation is often observed as the water 
grows colder in the late autumn. The flocks become very 
small and act imperfectly in entangling the suspended matter 
present in the water. Bacterial efficiency is decreased and 
aluminum hydroxide even passes into the filtrate. The addition 
of clay to the water after the alum dose has been well mixed 
has frequently, but not always, been found a remedy for " pin- 
point" coagulation. The quantity of clay employed will vary 
with the occasion, but a dose equal in amount to that of the 
alum is what should be used at the start and then lowered 
or raised according to the experimental results. The clay 
■should be added in fine suspension. 
An artificial schmutzdecke of previously precipitated alu- 
minum hydroxide could, of course, be applied to the sand surface, ARTIFICIAL PURIFICATION OF WATER 
179 
but at added cost not only for the extra chemicals, and labor, 
but also because of the decreased period between washings. 
"Pin-point" coagulation usually begins to be troublesome 
when the water temperature falls to about 400 F. 
Increased viscosity of water because of reduced temperature 
will have an appreciable effect upon the "loss of head" because 
of the increased resistance offered by the sand-bed. 
"Air binding" of the sand-bed is frequently an annoying 
outcome of winter weather. Cold water dissolves air more 
readily than that which is warm, and consequently a portion 
of that held in solution escapes when the water temperature 
rises in the filter-house. 
The bubbles of released air run together and finally burst 
through the upper sand layers, forming channel-ways which 
allow the passage of improperly filtered water. 
Frequent washing is a remedy, although a somewhat ex- 
pensive one. 
The "excess coagulation" method has been used for a number 
of years at Springfield, Mass., by Mr. E. E. Lochridge, and 
found to work well. The alum is applied in a stronger dose 
during six to seven hours and then shut off for the balance of 
the twenty-four hours' run. By such method of dosing good 
results are obtained by the use of less than half the total amount 
of alum required for the continuous feed. 
"The sulphate of alumina must be applied intermittently 
and at proper intervals. Thus, the same total amount of coag- 
ulant which gives excellent results by the method already 
described has no effect whatever upon the color if applied 
continuously through the twenty-four hours, or through twelve 
hours of the day." * 
The question so frequently asked, which system of filtration 
is it better to adopt, "slow-sand" or "mechanical" is not capable 
* Eng. News, Nov. 13, 1913. 180 
WATER-SUPPLY 
of an immediate answer. Local conditions must first be 
known. 
Either method of filtration, to be of value, requires con- 
stant attention. Careful watch must of necessity be kept of a 
mechanical plant, however small, to secure its running at all, 
and equal care should be given a slow-sand filter, although 
efficient supervision is seldom given the smaller beds. Once 
constructed, they are often left to look after themselves until 
cleaning is imperative, and then laborers are put upon the 
work who have never seen its kind before. 
It is very poor policy to build a slow-sand bed unless the 
plant be large enough to maintain a skilled force constantly 
employed. It may be said that, although relatively expensive 
for small plants, the slow-sand system rapidly lessens in cost, 
per unit of capacity, as the plant increases in size. This de- 
crease is not so rapid for mechanical filters. 
Other things being equal, slow-sand plants are suited for 
clear waters, and mechanical filters best for those of turbid 
or colored character. 
The life of a slow-sand filter is practically unlimited, while 
that of the mechanical type is as yet undetermined. 
Finally, it is an easier and cheaper matter to keep the 
mechanical filter nearer the present requirements as to capacity 
and to add to it by small extensions as circumstances may 
demand. So far as bacterial efficiency is concerned, either 
type will satisfactorily meet the requirements if carefully 
managed. 
Experiments of extended and costly character have been 
made with a view of securing data whereon to base a decision 
as to the merits of rival systems of purification, and it is quite 
reasonable that the question should be asked, Why is it nec- 
essary to multiply such tests? Why cannot Louisville and 
Pittsburgh results be made to do service elsewhere? To a large 
degree such use is possible, but it must be remembered that 
waters differ from each other more widely than the bulk of 
people appreciate. Thus the mere item of suspended matter 
is a case in point. Not only does the amount thereof greatly ARTIFICIAL PURIFICATION OF WATER 
181 
vary, but the changing character of the material causing the 
turbidity, and especially its degree of fineness, demands particu- 
lar attention. Two equally turbid waters may look alike to the 
eye, but the fine suspended matter of one may be easily arrested 
by a slow-sand bed, while the other may give a filtrate of 
decided opalescence. Hence the demand for local experiments 
in important cases. 
Nearly akin to the filter-plants referred to above is a group 
of appliances among which is the "Anderson process" a device 
well known in France. In this process the water is forced 
through iron cylinders revolving on hollow trunnions which 
serve for inlet and outlet pipes.* On the inner surface of 
the cylinders are curved ledges running lengthwise, which 
scoop up iron borings or punchings and shower them down 
into the water, as it flows through the cylinder. The water 
issuing from these "revolvers" is exposed to the air, whereby 
the dissolved iron is rendered insoluble, and by filtration through 
sand this precipitate is subsequently removed, together with 
what suspended matter it may have entangled. 
Good results are secured by the Anderson process at Boulogne 
and Choisy-le-Roi, but the use of the system involves an addi- 
tional and expensive step in the item of revolving machinery, 
before the water is run upon the sand-filters; and this further 
outlay of capital is not warranted, in view of the cheapness 
of some of the other and simpler methods of water improve- 
ment. 
A plan of the Boulogne installation is given on page 184. 
It consists of two "revolvers" f (4 feet 6| inches in diameter 
and 12 feet io| inches in length) through which the raw Seine 
water is pumped. The water takes three and a half minutes 
to pass through the "revolver," after which it is delivered to 
an "aerator" consisting of inclined troughs with step-like 
obstructions to break the current. The greater portion of the 
insoluble ferric compound formed in the "aerator" is permitted 
* See p. 182. 
t Only one is shown in the illustration. 182 
WATER-SUPPLY 
" revolver." Anderson's process. ARTIFICIAL PURIFICATION OF WATER 
183 
anderson's process at boulogne-sur-seine, 184 
WATER-SUPPLY 
to deposit in the "decanters" (i.e., long troughs in which 
the water passes alternately under and over the division- 
walls) and thence the water passes to the sand filter-beds. 
Each "revolver" has a capacity of treating 2500 cubic 
metres (660,000 U. S. gallons) of water in twenty-four hours. 
Four grammes (about 62 grains) of iron are required for each 
cubic metre (264 gallons ) of water. 
The "Drifting-sand" filter installed at Toronto is almost 
too new to allow of much being said with reference to its fitness 
for the work assigned it; but the published experimental results 
are good, and the plant looks promising, although the up-keep 
due to wear may be found to be expensive. 
The following description by Dr. G. G. Nasmith, of the 
Toronto Board of Health, is from the Canadian Engineer, 
April 8, 1915: 
"The filter consists of a vertical cylinder of reinforced 
concrete deeper than usual, in filter design, and is nearly filled 
with sand. The filtered-water collecting system is at the 
bottom divided into a number of small units consisting of a 
number of hoppers filled with graded gravel held in place by 
a brass screen fastened down to the concrete. At a distance 
of 2 feet 6 inches above the under-drain system and all around 
the filter are a series of hopper-shaped ports down which the 
drifting sand passes. These ports converge downwards to 
extractor outlets, with glass inspection tubes so that the opera- 
tion of the drifting sand can be watched. The extractors are 
controlled by cocks which can be adjusted to suit the char- 
acter of the water being filtered. 
" The drifting sand from the inspection outlets is collected 
by a ring main system along which is a current of wash-water 
which conveys the sand to the sand-washer or sand-separator. 
The sand-washer consists of a flat-sided hopper-like vessel 
with glass windows at the flat sides, through which the action 
can be observed. The sand-washer is attached to the filter 
supply-pipe immediately over a constriction in it, something 
like the form of a Venturi meter tube. An opening in the 
bottom of the sand-washer lets through the clean sand into the ARTIFICIAL PURIFICATION OF WATER 
185 
SECTIONAL VIEW OF THE DRIFTING-SAND FILTER. 186 
WATER-SUPPLY 
water passing, which conveys it to the filter and maintains the 
conical heap of sand in it. By the well-known hydraulic 
principle the pressure in the pipe at the constriction is very 
much less than on the filter. This difference is sufficient to 
draw the loose or drifting sand charged with dirt out of the filter 
into and through the sand-washer and into the supply-pipe, 
and so maintains the continuous motion of the sand while the 
filter is in operation. The sand, which by this time has become 
freed but not separated from the dirt, is led into the sand-washer 
at about mid-height. The weight of the sand carries it to the 
bottom of the washer and the dirty water escapes at the top. 
Near the bottom of the washer is a jet of water which carries 
the sand downwards into the supply-pipe, displacing the dirty 
water upwards in so doing. The water used for washing the 
drifting sand is unfiltered. The plan area of this filter is 143 
square feet, or 0.0033 of an acre. The estimated surface of 
the stationary cone is 306 square feet or more than twice the 
plan area. The average depth of the sand in the filter is about 
13 feet, of which about 54 per cent is drifting. The minimum 
thickness of drifting sand is about 2 feet 6 inches, and the 
maximum 10 feet. The minimum thickness of the stationary 
sand is 2 feet 2 inches and the maximum 13 feet. 
" The sand used was a local sand containing a large propor- 
tion of limestone, and was prepared principally by screening 
out the coarser materials and washing away the finer. The 
physical analysis of this sand shows its effective size to be .4 
m.m. and its uniform coefficient 2.7, and was passed through 
a sieve of eight meshes per lineal inch. In the drifting sand 
the coarser particles tend to collect in the planes of maximum 
travel and the finer materials to drop through the coarser, 
thereby offering the coarser materials to the water first and 
the finer materials later." 
The Puech-Chabal system of filtration is closely allied to some 
of the processes already described. It consists of a series of 
sand-beds progressively increasing in area and also in fineness 
of the sand grains. No coagulant is used. In a word, it is a ARTIFICIAL PURIFICATION OF WATER 
187 
refined "scrubbing" system, ending in a slow-sand bed. One 
of its advocates makes use of the following illustration: "If 
one desires to rinse out a beaker containing 1 c.c. of sulphuric 
acid, it is better to divide the wash-water into three or four 
portions and apply these in succession." A conception of the 
arrangement of the series of filters may be had from inspection 
of the following diagrammatic view and the cut on page 188, 
both taken from a reprint sent the writer by Mr. Walter Clemence, 
author of a paper on the plant at Madgeburg, Germany, pub- 
lished in Engineering, Jan., 1910. 
A number of rather large plants of this type are in opera- 
DIAGRAMMATIC VIEW OF PUE.CH-CHABAL FILTERS 
tion in Europe, but the system has no foothold in America, 
nor is there likelihood of its being adopted here. Mr. Walter 
Clemence, a strong advocate of the Puech-Chabal filter, writing 
to the author regarding the plant at Nanterre, France, says: 
"The capacity is io million Imperial gallons (12 million 
U. S. gallons) per twenty-four hours. The total cost of the 
plant is rather under £75,000 ($375,000)?' 
This is a high figure for a plant of that size, compared with 
mechanical filters as built here. As to efficiency, good results 
are secured, but at least as good can be obtained by the cheaper 
means at our disposal. 188 
WATER-SUPPLY 
The 11Non-submerged Filter" occasionally seen in France, 
especially at Chateau d'Un, is practically unknown to the 
American public. It is simply a "sprinkling filter" and re- 
minds one of the devices seen at sewage-disposal plants; differ- 
ing from them in having fine sand instead of broken stone for 
the bed and in having the water applied continuously in spray 
instead of intermittently. 
Although the results are good, the volume of output is low 
PUECH-CHABAL KILTERS, SHOWING CASCADES. MAGDEBURG, GERMANY. 
and the expense per million gallons filtered is consequently 
high. It can be considered only for a small supply. 
CHLORINATION 
The use of chlorine, either free or as a hypochlorite, is 
purely a matter of disinfection. Protection is offered against 
objectionable bacteria, but improvement in the physical or 
chemical character of the water is not contemplated. There- 
fore the process is commonly utilized as an adjunct to some ARTIFICIAL PURIFICATION OF WATER 
189 
previous treatment and is not very often used alone, except 
as a temporary expedient. Its employment began about 1909 
and has spread with rapidity, so that to-day it is the usual 
thing to find municipal water works either actively making use 
of chlorination or at least provided with means of quickly 
doing so upon emergency. 
" Bleach," so-called hypochlorite of lime, CaCloO, is the 
form in which chlorine is most frequently administered. The 
sodium salt NaClO, formed by electrolyzing a common salt 
solution, has been used, but not as widely as the calcium com- 
pound. The manner of using " bleach " is to extract it with 
water and then add the more or less clear liquid to the water 
supply immediately before distribution to the consumers. If 
filtration be a part of the water treatment, the "bleach" solu- 
tion, or "hypo," as it is commonly called, should be added in 
the clear-water basin, that is, after filtration is accomplished. 
This is important, for the reason that color and suspended 
matter use up a large quantity of the chemical and heavy 
doses are required to secure satisfactory bacterial results in 
colored or turbid waters. 
Stock solutions of the "bleach" are made up of a strength 
varying in different places from one-half to two per cent; these 
are stored in suitable tanks, and are tapped therefrom through 
accurately standardized orifice boxes. Except for emergency 
plants, which are built of wood, the " bleach-tanks " should be 
of concrete to avoid corrosive action. 
The dose of "bleach" required is stated in parts per million 
of "available chlorine," for the reason that although the chem- 
ical is supposed to contain about one-third of its weight as 
chlorine, yet it will vary in strength with its age and method 
of packing, thus requiring frequent analysis. Stated in the 
form above, the ordinary dose for clear water is about 0.35 
p. p. m. "available chlorine." For colored or turbid waters 
the dose would have to be increased. 
Overdose of bleach is to be carefully avoided, as its peculiar 
taste and odor will be noticed and complained of by the con- 
sumers. The sense of taste is more delicate than that of smell 190 
WATER-SUPPLY 
for its detection. People differ as to their ability to notice its 
presence, but a dose of from 0.5 to 0.6 p. p. m. available 
chlorine will be tasted by some persons, while nearly twice 
that quantity will be required to enable them to detect the 
odor. 
This is on the assumption that the water is clear, for, as 
has been said, heavy doses of bleach are used up by strong 
color or turbidity. 
Removal of the chlorine taste may be accomplished by the 
subsequent addition to the water of a solution of sodium thio- 
sulphate (Na2SaO3), the amount used being a weight of the 
crystallized salt equal to that of the available chlorine employed. 
This quantity will be in slight excess of the theoretical. Some 
water authorities believe in even extending the above dose to 
"half of the quantity of chlorinated lime to be on the safe 
side at all times and to take care of all reasonable overdosing 
with chlorine." * 
The occasion for the use of the thiosulphate as above, may 
arise from the accidental addition of an overdose of chlorine 
as well as from the intentional employment of it in large doses 
because of the excessive pollution of the raw water, as in times 
of emergency. 
It has been occasionally reported that the use of bleach 
tends to increase the odor naturally following the presence 
of odor-producing algce. 
Liquid, chlorine is displacing "bleach" in many plants. 
It ordinarily costs about eight cents per pound, while at 
normal times "bleach" can be had for a little less than 
two cents; but as the latter contains only one-third its 
weight of chlorine the difference in cost is really not great, 
and ease of administering an accurately adjusted dose of the 
liquid chlorine makes up for the added expense. Both of 
these chemicals are now much higher in price, due to 
the European war. 
* Engineering News, 69: 937. ARTIFICIAL PURIFICATION OF WATER 
191 
The dry, liquefied chlorine is purchased in steel cylinders 
under a pressure which varies from 54 pounds per square inch 
chlorine dosing apparatus. (Wallace and Tiernan.) 
A-chlorine tank. 
B-valve on chlorine tank. 
C-auxiliary valve 
D-flexible connection. 
E-pressure compensating valve for taking 
care of the varying pressures in the chlor- 
ine tank and also maintaining a constant 
drop in pressure across the valve S. 
5-control valve. 
G-valve to prevent moisture from getting 
back into control parts of apparatus. 
H-valve to control flow of water and to 
keep chlorine out of incoming water con- 
nections. 
F-pressure gauge showing pressure in 
tank. 
J-chlorine flow meter (inverted siphon 
type). 
I-chlorine absorption chamber. 
K-L-chlorine solution line. 
V-water connection. 
P-water valve. 
R-gauge. 
O-water valve to water seal N. 
T-water spill. 
U-chlorine solution line to point of 
application. 
N-water seal. 
X-water pressure reducing valve. 
Q- Y-strainers. 
Size of apparatus mounted in wall cabinet 
20X21 inches. 
at 320 F. to 216 pounds at 1220 F. and is added to the water 
by the employment of one of the several automatic devices 192 
WATER-SUPPLY 
now on the market, or else by some less ambitious apparatus 
contrived by the local management. The chlorine dosing 
apparatus here pictured is widely used and is efficient. The 
method of its working will be understood from the lettered 
description. 
Distribution through the water is conveniently accomplished 
by discharging through a disk of carborundum or similar porous 
material, which delivers the gas in very finely divided streams 
of bubbles. 
In large plants the labor of preparing the solution of " bleach" 
and seeing to its proper admission to the water is scarcely felt 
by the staff of well-trained attendants, nor is the analysis 
of various lots of the chemical a burden of consequence for the 
local laboratory; but in small establishments, with but few 
employees, this work may well be a serious matter, and in 
such places the change to liquid chlorine becomes a relief, 
even at a somewhat greater cost. 
The action of chlorine upon bacteria differs greatly with the 
character of the organism, thus bacillus tuberculosis is very 
resistant and the spores of all bacterial forms are practically 
unattacked except by excessive doses. Naturally such bacteria 
as are imbedded in organic material are protected from the 
action of the disinfectant. 
In this connection Winslow says: 
"We recognize three types of resistance to such oxidizing 
action as that of chlorine-ordinary vegetative cells, cells with 
a waxy envelope or of waxy substance, like the tubercle bacillus, 
and spores like anthrax and B. sporogenes, which are still more 
resistant to the action of the disinfectant. There is danger, 
therefore, that when the colon bacilli are removed these other 
types may remain." 
A. B. Gillie of this laboratory reports the following results 
of dosing clear water, artificially seeded with ordinary bacteria, 
with "bleach" to the extent of o. 5 part per million of available 
chlorine. 
Intervals of time between dosing with the "bleach" and 
"sowing" the water for the bacterial count are indicated. 193 
ARTIFICIAL PURIFICATION OF WATER 
After o minutes 220,000 bacteria per c.c. 
" 5 " 1,200 11 11 
" 30 " 800 11 11 
" 60 " 15 
t ( , „„ Cl „ c c c c 
120 2 
Similar data are given by E. J. Tully for the clear water of 
Lake Mendota, Wis.* The interval between dosing and sowing 
was thirty minutes. 
Bacteria per c.c. 
chlorine, p.p.m, 1 
0.0 8750 
0.1 2550 
0.3 4 
0.5 2 
1.0. 1 
There is no question about chlorine being able to greatly 
reduce the count of bacteria per cubic centimetre, but although 
we are interested in this reduction of the general count, we 
are much more concerned about the destruction of the patho- 
genic organisms. One German authority says: "It (chlorine) 
effects their complete destruction when used in a quantity 
which will not prevent subsequent putrefaction by harmless 
species"; f and that seems to be a practical if not an absolute 
truth. 
Some water specialists have looked upon chlorine as a 
"dope" for bacteria rather than as a germicide; claiming that 
its action is but temporary, and Dr. Hale and other bacteriol- 
ogists have found that "attenuation only of B. coli instead of 
destruction is frequent." J 
However true it may occasionally be that an apparent 
after-growth follows the use of chlorine, the fact remains that 
experiments upon a huge scale, dealing with great centers of 
* Amer. J. Pub. Health, May, 1914. 
t Wasser und Abwasser, i: 209. 
J Proc. N. J. San. Asso., Dec., 1914. 194 
WATER-SUPPLY 
population, have shown that, from a practical standpoint, 
chlorination has stood between the people and disaster. 
A very few instances, out of many that might be cited, are 
here given: 
At Erie, Pa., chlorination was adopted March 15, 1911, dur- 
ing a typhoid epidemic due to the pollution of the public water. 
The beneficial results were at once apparent. 
Mr. C. A. Jennings reports the typhoid rate for Kansas 
City for the years 1900-1910 inclusive as 42.5 per 100,000. 
Chlorination was introduced in January, 1911. The average 
typhoid rate for 1911, 1912, and 1913, was 20. 
Dr. G. G. Nasmith of Toronto writes: 
"Last year our intake plugged with sand and we were 
forced to short-circuit our water supply and take it from the 
bay into which all our sewage empties. We had the makings 
of the largest typhoid epidemic even known on this continent, 
but fortunately we had hypochlorite to depend on and we came 
through with a typhoid death-rate of 20 per 100,000 in 1911 
as compared with 45 per 100,000 in 1910, an extremely satis- 
factory showing." 
The cost of chlorination, if liquid chlorine be used, is prac- 
tically nothing beyond the price paid for the chemical (about 
eight cents per pound) after the dosing machine is once estab- 
lished. 
If "bleach" be employed the expense for labor is greater 
but still the gross sum is small. The cost of the chemical is 
ordinarily a little less than two cents per pound. 
Mr. J. S. Dunworthy estimates that at Erie, Pa., the 
"average cost of operating the sterilization plant together with 
the operation and maintenance of the laboratory is approxi- 
mately $1.00 per million gallons of water treated." 
The plant above referred to was one erected to meet an 
emergency and has now been replaced by a very complete 
mechanical filter plant with chlorination apparatus added. 
The cost of maintaining such disinfecting addition to a 
filter plant is practically nothing beyond the cost of the 
chemicals. 195 
ARTIFICIAL PURIFICATION OF WATER 
Swimming pools are now so widely in use that it might not 
be out of place to say that some care should be taken to make 
their waters as nearly potable as possible. Many of the bathers 
swallow more or less of the water accidentally, and some of 
them intentionally do so. The pools are of two kinds, those 
operated on the "fill and draw" plan, where the entire body 
of water is wasted after an interval and the basin is then re- 
filled, and those where the water is changed only after long 
periods, but is continually pumped through a filtration plant 
using alum, or through some other kind of purification apparatus. 
The pool at the Rensselaer Polytechnic Institute (100,000 
gallons capacity) is of the latter type, the water being con- 
tinually circulated through three very complete filters. 11 Bleach" 
is added in dry form to the surface of the water every week, 
the dose being 1 part per million of available chlorine. The 
result is satisfactory. 
OZONE 
Ozone, an allotropic form of oxygen, is an active agent, 
both chemically and as a germicide, because of the readiness 
with which it furnishes atomic or nascent oxygen. There are 
sundry ways in which it may be prepared, but the one of im- 
portance here is that by which electric discharges are passed 
through oxygen or atmospheric air. Because of that method 
of preparation, ozone strongly appeals to the public as a safe 
and proper agent for water purification; and such it would 
be were it not for the difficulty of securing intimate contact 
between the ozone and the water to be treated, for the expense 
attending its use, and for the further reasons that it does not 
remove color with certainty, and its effect upon turbidity is 
slight. It cannot, therefore, be used alone, except upon a clear 
and colorless water. 
To a certain degree, however, color and turbidity are altered 
by ozone. It is a strong oxidizer, hence some bleaching naturally 
follows its use and there may be a slight lessening of turbidity, 
due to action on suspended organic particles. 
Expense is the serious objection to the employment of 196 
WATER-SUPPLY 
OZ NE STERILIZING TOWER NIEUWERSLUIS, HOLLAND. ARTIFICIAL PURIFICATION OF WATER 
197 
this gas; not only because of the first cost of apparatus and 
the power necessary to produce the necessary electricity; but 
also on account of the expensive up-keep and the high-priced 
labor required. 
Stated tersely, the ozone process consists in "ozonizing" 
Separator 
DIAGRAMMATIC ARRANGEMENT OF THE DE FRISE SYSTEM AT ST. MAUR, PARIS. 
(From the paper of Dr. Samuel Rideal before the Royal Sanitary Institute.) 
atmospheric air by means of a silent electric discharge and then 
bringing such ozonized air into intimate contact with the 
water to be treated. It would seem a simple matter to accom- 
plish the second step, but ozone is only slightly soluble in water, 
so that intimate contact has to be accomplished by distributing 
fine bubbles of the gas through the water in a very thorough 
manner. 198 
WATER-SUPPLY 
Mixing the ozone and water is secured in the De Frise system 
as shown by the diagrammatic illustration on page 197, the 
water and ozone enter at the same point and mingle in the 
sterilizing tower, while in the Vosmaer apparatus the water 
flows downward through the tower and the ozone starts at 
the bottom and flows upward. Such a tower at Nieuwersluis, 
as seen by the writer (see p. 196), was of glass, but that was 
manifestly only for exhibition purposes. 
THE HOWARD-BRIDGE MIXER. 
Raw water enters the pipe a5 drawing, by suction, unabsorbed ozone from 53. 
The waste gases escape at Z>4. The current of water in a sucks fresh ozone 
into itself from the ozonizer c; and after passing around the baffle plates in 
b, the purified water escapes at b2. 
The Howard-Bridge mixer, page 198, is another device built 
with a view to secure good contact between the ozone and 
water. 
The air required for the ozone production should be dry, 
but not completely so, inasmuch as a very small quantity of 
aqueous vapor is beneficial, according to the researches of 
Arrhenius. This drying was formerly accomplished by the 
use of some drying agent, such as sulphuric acid or calcium 
chloride, but is now secured by refrigeration. 
The devices employed for the ozonization of air and the 199 
ARTIFICIAL PURIFICATION OF WATER 
means of administering it to the water are too many, and 
their description too voluminous, for insertion here. Reference 
must be had to the monographs on the subject.* 
Plants of large size have not been erected in America, so 
that opportunity has not been offered here for large-scale 
study of the process. The plants at Ann Arbor, Mich., and 
Lindsay, Ont., have been discontinued. A considerable num- 
ber have been installed in Europe and have been examined 
while working by such men as Rideal, Gerard, and others. 
Rideal found the De Frise system at the Paris waterworks 
Filling in.D and B Open,A and.C Closed, 
Filtering A-Az" 'D «« B " C « 
Cleaning A " C " A-B <« D " 
ARRANGEMENT OF THE MARTINIKENFELD PLANT. 
(From report of Dr. J. J. de Kinder.) 
of St. Maur giving very successful results. Water from the 
River Marne is used after sedimentation and filtration. After 
ozonizing it is a "well-aerated brilliant liquid of natural 
bluish tint and quite inodorous," containing but one or two 
innocuous organisms and no B. coli. The absence of B. coli, 
and pathogenic organisms as well, in ozone-treated water has 
been frequently observed, and the expression "selective" has 
been often used to describe the action of the gas upon that 
type of bacteria. For instance, the following figures are given 
by the United Water Improvement Co. It will be observed 
* See in this connection a full article in Engineering News, 63: 488. 200 
WATER-SUPPLY 
that the effect upon "color" is variable and that the action 
upon "turbidity" is nil. In all cases the character of the 
substances producing color or turbidity will alter the effect 
of the gas. 
TABLE SHOWING THE EFFECT OF OZONE ON WATER 
Bacteria per c.c. 
Coli. 
Color. 
Turbidity. 
Raw. 
Ozon- 
ized. 
% Re- 
moved 
Raw. 
Ozon- 
ized. 
Raw. 
Ozon- 
ized. 
% Re- 
moved. 
Raw. 
Ozon- 
ized. 
% Re- 
moved. 
2,460 
26 
98.95 
Present 
Not found 
22 
5 
77.28 
20 
20 
00.0 
2,800 
12 
99.58 
30 
15 
50.00 
40 
40 
00.0 
1,100 
5 
99-95 
17 
0 
100.00 
15 
15 
00.0 
1,270 
6 
99-53 
35 
15 
57.15 
15 
15 
00.0 
1,200 
0 
100.00 
12 
0 
100.00 
17 
17 
00.0 
59,000 
680 
98.85 
65 
40 
38.47 
400 
400 
00.0 
Ozone is not only destructive of bacterial life, but, in the 
quantity required for water improvement, is harmless to life 
of the higher orders. Moreover, by its very activity, it is 
self-destroying, and any excess is quickly disposed of. These 
characteristics are strong recommendations for its use in the 
betterment of water. 
The ozone concentration is a term signifying the number 
of grams of ozone contained in one cubic metre of ozonized 
air. The following table, prepared by the authorities in charge 
of the examination of the De Frise plant at St. Maur, gives 
the concentration there used together with sundry other 
data.* 
Concentration 
Grams of Ozone 
per Cubic Metre 
Air. 
Cubic Metres of 
Air Used per 
Cubic Metre of 
Water. 
Total Ozone Used 
in Grams per 
Cubic Metre of 
Water. 
Bacteria per c.c. 
Before Ozoniza- 
tion. 
Bacteria per c.c. 
After Ozoniza- 
tion. 
i .6 
I 383 
2.216 
800 
2 
i .04 
2-035 
850 
4 
.98 
2.077 
1-352 
2682 
3 
.88 
i-537 
2.04 
3732 
7 
1. 
1.003 
I. 
2886 
9 
•835 
1-003 
0.693 
2886 
45 
1.12 
' 1 037 
1.61 
4445 
14 
i-43 
0.831 
1.188 
145 
2 
* See J. Franklin Institute, May, 1907. ARTIFICIAL PURIFICATION OF WATER 
201 
This "dose'' of ozone required at St. Maur is given in the 
third column of the above table. The amount of this dose 
will vary somewhat with different waters; thus at Brussels, 
Gerard secured good results by the use of 0.3 gram per cubic 
metre (264 gallons) of water. He employed sterilizing towers 
five metres high arranged with multiple diaphragms to increase 
the bubbling of the ozonized air. As Chassy has shown, it 
is much cheaper to employ a large volume of feebly ozonized 
air rather than a smaller volume with a high concentration. 
"To obtain a concentration of seven per cent, ninety times as 
much energy is required as for a concentration of one-half of 
one per cent." 
The cost of ozone treatment is a very vexed question that 
has not yet reached a satisfactory solution. The process is 
in competition with the sundry methods of filtration and with 
sterilization by chlorine. There is no question about its being 
greatly higher in price than the latter, although it certainly 
possesses the advantage of presenting a more advantageous 
system of procedure, considered from the popular viewpoint. As 
compared with filtration, the question naturally arises, inasmuch 
as clarification of the water is usually required before ozone can 
be applied, why not increase the efficiency of the filter plant, 
at a slightly greater expense, and omit the ozone altogether ? 
Mr. I. M. de Varona conducted work with the experimental 
plant erected for the City of New York and found the cost 
amounted to the prohibitive figure of about twenty dollars 
per million gallons of water treated. These figures were vio- 
lently attacked by the friends of the process. One of them, 
Mr. Bridge, referred to the results secured by Dr. Erlwein in 
Germany, where, using 1.3 gram of ozone per cubic metre 
of water treated, the cost was one-half a cent per cubic metre 
or $18.92 per million gallons, using an experimental plant, 
and an estimated cost as low as $7 per million gallons, if a 
plant of ten million gallons capacity were employed.* 
Prof. Leon Gerard of the University of Brussels made a 
* J. Franklin Inst., May 1907. 202 
WATER-SUPPLY 
full report upon the ozone process plant at Breda. He fond 
that the cost for the ozone treatment, exclusive of any prelim- 
inary treatment or interest charges, was $8.86 per million 
U. S. gallons.* 
Miquel reports to the Paris Municipal Council, that "the 
process has been found capable of eliminating a large propor- 
tion of the bacteria present in water of various descriptions 
and of permanently destroying with certainty the germs of 
bacillus coli, which possess greater resisting power than the 
Eberth bacillus and the cholera spirillum. The cost of this 
treatment is stated, in dealing with large volumes, to be about 
centimes per cubic metre, say $15 per million gallons.' 
S. T. Powell gives the following data for the Herring Run 
plant near Baltimore, which is a ten-million gallon unit under 
his observation. 
With an ozone concentration of 1.05 (i.e., 1.05 grams 
ozone per cubic metre of air) the bacteria per c.c. were reduced 
from 2720 to 25, a percentage removed of 99.1. 
He places the installation cost at $5000 per million gallons 
capacity and the maintenance at $4 per million gallons de- 
livered, assuming interest at 5 per cent, depreciation at 5 per 
cent, and electric current at 3 cents per kw. hour.f 
ULTRA-VIOLET LIGHT 
Ultra-violet light is the latest contribution to the list of 
processes for water purification upon a considerable scale. 
It consists, in a word, of the exposure of the water in thin 
layers, and at short range, to the light of an electric mercury 
vapor lamp with walls of quartz. Usually provision is made 
for more than one exposure of the same water to the light. 
The illustrations which follow are obtained through the 
courtesy of the R. U. V. Company. 
"In this apparatus the lamp is enclosed in a box, through 
which the clarified water from the roughing filters runs. This 
*La Technologic Sanitaire, March 15, 1905. 
t J. N. E. Water Works Asso., 29:87. ARTIFICIAL PURIFICATION OF WATER 
203 
box is provided with baffle plates, which cause the water to 
pass by the lamp in the center more than once, so that 
there is assurance that every particle of the water will be within 
the range of the ultra-violet rays some time during its passage 
through the circular box. The lamp itself is separated from 
the water by windows of quartz glass, which has the property 
ULTRA-VIOLET RAY APPARATUS. 
of absorbing only a negligible minimum of the ultra-violet rays 
thrown out from the mercury arc lamp." While quartz is 
transparent to these rays, ordinary glass is opaque. 
For plants of large capacity a canal installation is used such 
as shown herewith: 
This system of water purification resembles the ozone 
method in that it requires the water to be clear and colorless 204 
WATER-SUPPLY 
before the light can be efficiently applied. It is more neces- 
sary to remove color and turbidity before using the ultra- 
violet ray than it is to insure their absence when about to 
employ ozone. The ray does not penetrate far even through 
clear water. 
The cost of sterilizing with ultra-violet light will depend 
upon the local price for electric current, together with interest 
STERILIZING CANAL WITH TEN LAMPS IN SERVICE IN THE CITY WATERWORKS OF 
LUNE VILLE, FRANCE. 
The lamps here employed are of the " pistol " type, and are inserted opposite baffles in 
the canal. See illustration on next page. 
charges for the first cost of the plant. This latter item will 
depend upon the type of pre-filters erected to remove color and 
turbidity from the raw water. The staff required to manage 
the filters will be enough to look after the sterilizers as 
well. Inasmuch as some form of filter is usually a necessity, 
the question arises, as in the case of ozone treatment, why 
not put in a high-grade filter, and dispense with the sub- 
sequent portion of the installation? Further development ARTIFICIAL PURIFICATION OF WATER 
205 
and experience will doubtless answer this question for both 
processes. 
G. Espitallier,* as a result of investigation, estimated that 
sterilization could be secured by the employment of 26 watt- 
hours per cubic metre of water treated, i.e., 264 U. S. gallons. 
Recklinghausen gives a somewhat larger figure, namely 36 
watt-hours per cubic metre of water. The water contained 
5000 B. coli per c.c.f He also finds that bacteria differ in 
sensitiveness to the ray, the killing of sundry forms taking place 
in the following periods of exposure: 
VIEW OF PISTOL LAMP. 
Staphylococcus 5 to 10 seconds 
Cholera 10 " 15 " 
Coli ' 15 " 20 " 
Typhoid 10 " 20 " 
Dysentery 10 " 20 " 
Subtilis 3°" 60 " 
Tetanus 20 " 60 " 
E. A. Mau of this laboratory found the influence of color 
in the water to be slight so far as diminishing the action of 
* Eau et Hygiene, Jan. 1911. 
f Engineering Record, 64: 60. 206 
WATER-SUPPLY 
the ultra-violet ray was concerned. Using extract of dead 
leaves as a coloring liquid, waters heavily dosed therewith 
were sterilized by the light, although the time of exposure 
had to be somewhat extended. 
The interference due to turbidity was noted to be more 
marked than that of color and it was discovered to be the same 
in degree irrespective of whether the turbidity was caused by 
organic or inorganic material. Increased velocity of the water 
was shown to diminish the sterilizing effect of the light, as 
was to have been expected, because of the resulting shortening 
of the time of exposure; but the interesting observation was 
made that a critical velocity existed up to which beneficial 
results were secured, doubtless for the reason that the eddy 
currents set up by the quicker flow carried deeper layers of 
water more certainly within range of the light. This good 
effect became more than balanced later as the rate of flow 
became more swift. Irregularities in the bottom and sides 
of the flow-channels such as would produce eddy currents, are 
therefore indicated as of practical benefit. 
From the sanitary standpoint, the method of purification 
of water which excels all others in efficiency is distillation.* 
The peculiar taste of freshly distilled water is, however, 
disagreeable to many, and for that reason the process is not 
likely to become speedily popular, even if the expense-were not 
too great. Some years ago, in a paper before the American 
Society of Civil Engineers, Mr. Hill advocated the use of dis- 
tilled water for city drinking supply, and basing his calculation 
upon one million gallons daily, delivered by small separate 
distribution-mains, he estimated that the total cost of the 
water, interest charges included, would be one-eighth of a cent 
per gallon, which would be at the rate of $1250 per million 
gallons. High as this price at first appears, it is nevertheless 
much cheaper than what is now paid by many people for bottled 
11 spring water." 
* Further consideration of this question was had on p. 114. ARTIFICIAL PURIFICATION OF WATER 
207 
Distilled water is used for drinking purposes on practically 
every vessel in the United States Navy, and former Surgeon- 
General Tryon wrote: "It may be stated that the medical 
officers of the navy recognize the great value of distilled water 
in the improvement in health that has followed its intro- 
baird's condenser. 
duction, particularly on certain foreign stations." That the 
present Surgeon-General of the Navy also entertains a high 
opinion of its value we have already seen. 
The apparatus used upon the vessels is one devised by 
Chief Engineer G. W. Baird, and called by his name. An 
especial feature of value is the introduction of the steam into 208 
WATER-SUPPLY 
the condenser in such a manner as to drag with it a constant 
and regulated current of air, thereby causing efficient aeration 
during the very act of condensation. By this means and the 
subsequent passage through a bone-black filter, the water loses 
much of the disagreeable taste above referred to, and by fur- 
ther standing for some twelve hours the taste entirely dis- 
appears. 
That complete sterilization takes place at the high tem- 
perature attained there can be, of course, no question. Even 
the spores of the now known pathogenic bacteria are rapidly 
destroyed by exposure to the temperature of boiling water, 
although those of certain non-pathogenic varieties will resist 
that temperature for hours. 
Some years ago a large multiple-effect distilling plant was 
erected at Fort Jefferson, on the Dry Tortugas, Fla., to furnish 
water for the naval station and barracks at that place, and to 
such vessels of the navy as may make port at that point in 
need of fresh-water supplies.* 
A further illustration of a large distilling-plant may be 
seen at Gibraltar. The water supply of that town and fortress 
is normally stored rain-water, but the authorities are always 
prepared to furnish distilled sea-water in case of need. 
Aeration of water has always held in the public mind a 
position of prime importance as a means of purification. There 
is unquestioned benefit arising from the passage of water over 
falls and rapids, and from its being thrown into the air in 
fountains, but the benefit is not so great as is commonly 
believed. 
Agitation and aeration tend to prevent the abundant growth 
of algae, with their objectionable tastes and smells, and helps 
to shake out volatile or gaseous products of ill-smelling character, 
so that undoubted improvement in quality of water results 
from the establishing of a fountain in, or otherwise blowing 
air into, a too quiet reservoir; but the expectations of those 
* See Engineering News, March 20, 1900. ARTIFICIAL PURIFICATION OF WATER 
209 
AERATOR AT WEST PARISH FILTERS, SPRINGFIELD, MASS. 
(From report of Water Board.) 210 
WATER-SUPPLY 
who hope to thus easily eliminate pollution of a more serious 
character will not be realized. 
A further consideration of the influence of aeration will 
be found in the chapter on stored water. 
Iron removal by the use of aeration and subsequent filtration 
is resorted to with ferruginous waters which are otherwise fit for 
general domestic use. As a result of wide inquiry among those 
interested in water for laundry purposes the writer concludes 
that more than 0.5 part per million of iron (stated as Fe) is 
unacceptable for such use. Objection is occasionally raised 
to a smaller amount. 
By blowing air into such waters, spraying them or even by 
letting them stand freely exposed to the atmosphere, carbon 
dioxide is reduced and the iron is commonly (but not always) 
oxidized to an insoluble form easily removed by filtration. 
Coke filters, or tricklers, are frequently used to accomplish 
the aeration and oxidation needed and it has been found that the 
metallic iron in the coke increases the efficiency of the filter 
for removal of the iron in solution.* 
The simple method of iron removal by aeration is indicated 
when the iron is present in an easily oxidized form, such as the 
carbonate. The process may be seen in operation at Asbury 
Park, N. J., and also at Far Rockaway, N. Y. 
At the former plant the water is derived from deep wells 
(400 to 1100 feet deep) and is raised by an " air-lift." After 
reaching the surface it is forced through pressure filters of the 
ordinary mechanical type. Alum is not used. At Far Rock- 
away the principle is the same, only an " air-lift " is not required. 
Sufficient oxidation is secured by allowing the water to spill 
from the upturned end of the pump-main, after which it is 
run upon two slow filter-beds which are open and of the 
English type. In each of the above cases the result is satis- 
factory.! 
Even if the iron be fully oxidized, it may fail to precipitate, 
* J. N. E. Water-works Asso., xi, 278. 
f Engineering News, June 4, 1896, and April 12, 1900. ARTIFICIAL PURIFICATION OF WATER 
211 
because of there being considerable organic matter present 
in the water tending to hold it in solution. 
Downs reported that the iron in a colored water of the 
canal zone was in union with organic matter, and utterly failed 
to be removed by aeration. He was successful in his efforts 
to take it out by the use of a dose of 1.5 grains per gallon of 
aluminum sulphate, followed by sedimentation and filtra- 
tion.* 
Should the iron be present in the form of ferrous sulphate, 
then its removal becomes somewhat less easy. 
Sulphate of iron has been successfully removed from the water 
of Reading, Mass., by the use of lime, followed by aeration and 
mechanical filtration,f but studies are now under way by 
R. S. Weston with a view to abandon the treatment with 
chemicals and resort to aeration and the contact-action noted 
further on. 
In America a number of iron removal plants are in operation, 
but space can be spared for a brief description of only two 
beyond those already mentioned. 
At Middleboro, Mass., the water from the well is pumped 
to a " trickier " built of concrete, of cylindrical form, thirty 
feet in diameter and containing a bed of coarse coke ten feet 
in thickness. The entering water is sprayed upon the coke 
and after passing the bed, flows to a settling basin and thence 
to the two filters. Arrangement is made for a yearly flushing 
out of any excess of iron deposit on the coke by filling the 
" trickier " with water and then allowing it to suddenly flow 
to waste. 
The two compartments of the filter have a combined sand 
area of 0.1 acre. The beds are 12 inches of graded gravel 
supporting 36 inches of sand of 0.31 m.m. effective size. A 
full description of this plant is given by Professor Weston, 
J. N. E. Water-works Asso., 28: 27. 
The capacity of the plant is one million gallons per day. 
It cost $18,000. 
* Proc. Am. Water Works Asso., 1910. 
t Engineering News, June 4th and Nov. 26th, 1896. 212 
WATER-SUPPLY 
COKE TRICKLER FOR IRON REMOVAL, MIDDLEBORO, MASS. 
COKE TRICKLER FOR IRON REMOVAL, MIDDLEBORO, MASS. (iN OPERATION) ARTIFICIAL PURIFICATION OF WATER 
213 
AVERAGE RESULTS OF OPERATION 
(Parts per million) 
Determination. 
Well. 
Settling Basin 
Effluent. 
Filter Effluent. 
Color 
48.0 
22.0 
5 •0 
Turbidity 
5° 
- 3-° 
1.0 
Iron 
1.62 
0.46 
0.17 
Manganese 
0.67 
0.36 
0.27 
Hardness 
27.3 
23 -4 
Oxygen consumed 
i-79 
1-53 
I 05 
Carbon dioxide 
41.0 
4.2 
4.6 
Dissolved oxygen 
2-95 
10.20 
9-55 
The Middleboro water shows a tendency to increase in the 
amount of contained iron as is indicated by the following figures 
given by the Mass. State Board of Health: 
1893, 0.7; 1894, 0.237; 1895, 0.187; 1896, 0.288; 1897, 
0.227; 1898, 0.408; 1899, 0.329; 1900, 0.489; 1901, 0.487; 
1902, 0.841; 1903, 0.922; 1904, 1.372; 1905, 1.17; 1906, 1.126; 
1907, 1.209; 1908, 1.349 parts per million. 
This is a condition often observed, especially in districts 
where the ground-water has been " over-pumped," and it is a 
discouraging one if an iron-removal plant be not at hand. 
With such a plant available, however, the increase in dissolved 
iron is unimportant, and may be even beneficial, inasmuch as a 
reasonably large amount of iron is easier to remove than a small 
quantity. 
The plant at Lowell, Mass., designed by F. A. Barbour, 
possesses a coke bed ten feet deep, an intermediate sedimenta- 
tion basin of one hour capacity and a sand filter three feet deep; 
but the point in which it differs from the Middleboro plant 
is the way in which the coke bed is used. In the latter city it 
is employed as a sprinkling or trickling filter, the water being 
sprayed over its surface as already described; while at Lowell 
the coke is used as a " contact-bed," much after the fashion 
of the sewage treatment-device bearing the same name. 
The surface of the coke is not flooded by the applied water, 
nor are any pools formed thereupon; but the filter is " back 
flooded " to within a short distance of the coke surface and the 214 
WATER-SUPPLY 
water is applied in streams about one inch thick, so that there 
is much " contact " and only partial aeration by spraying. 
The amount of both " contact " and aeration can be adjusted at 
will. The size of coke is about two inches. 
Mr. Barbour has shown that the iron and manganese present 
in the Lowell well water cannot be removed by aeration and 
direct application to sand filters; nor by aeration, sedimenta- 
tion and sand filters; and that excessive aeration by a sprink- 
ling filter is not adapted to the treatment of the water; but 
that a coke bed, used as a contact-device in the manner noted 
above, followed by sand filtration, will remove the iron and 
manganese successfully and economically, at a cost of $7.65 
per million gallons, including interest, depreciation and oper- 
ating charges. See his report to the Municipal Council, 1914. 
Manganese in water will cause a troublesome turbidity upon 
exposure to air and will make the associated iron somewhat 
more difficult of removal. 
The method of treating such a water is similar to that 
employed for one containing iron alone as has been shown, 
but the separation is slower and therefore the contact-action 
should be more prolonged. 
Reference is made on page 485 to the Permutit method for 
manganese removal. 
Carbon dioxide removal from water is required if it be 
present in considerable quantity, for the reason that the gas will 
strongly attack lead pipe and lead to " plumbism " or lead 
poisoning, and it will act as a corrosive when the water is used 
in boilers. 
Lime is the chemical agent that naturally suggests itself 
but the fact that its use hardens the water is a disadvantage. 
The amount that Barbour was obliged to use at Lowell (1.5 
grains per gallon) so increased the hardness that he found it 
more expedient to displace the dissolved gas by aeration. 
He employed small nozzles | to | inch in diameter which 
threw a very fine spray and caused a very effective removal of ARTIFICIAL PURIFICATION OF WATER 
215 
the carbon dioxide, reducing it from 45 parts to 3.3 parts per 
million. The pressure at the nozzle was five pounds per square 
inch. The cost did not exceed $2.25 per million gallons, in- 
cluding interest, depreciation and pumping. 
This cost is less than that of the lime treatment, if considera- 
tion be taken of the expense due to the extra soap required to 
meet the added hardness. 
Household filtration on the domestic scale is in very general 
operation, yet satisfactory results are obtained in but a small 
percentage of cases. 
The companies manufacturing the mechanical filters already 
mentioned make sizes intended for domestic use, but the 
skilled labor furnished by a city employee whose sole duty 
it is to attend to the public plant is very rarely obtainable in 
the average household; consequently the filter is neglected or 
mismanaged, or both. In short, filtration, to be effectual, 
should be municipal. A house filter that has come widely 
into use, and upon which very many people pin their faith, is 
the well-known " Pasteur." It is commonly operated under 
the pressure of the city mains, but may also be arranged to 
work without additional pressure beyond that of the atmos- 
phere. The cut herewith given shows its simplest form, and, 
for those unacquainted with its use, it may be said that it con- 
sists of a cylinder of fine unglazed porcelain (called " candle " 
on account of its size and shape) enclosed by one of metal; 
and that, connection having been established between the 
latter and the supply-pipe, the water is forced through the 
pores of the porcelain to the inside of the cylinder (the so- 
called candle), whence it drops into the reservoir, leaving the 
suspended matter as a coating upon the 11 candle's " exterior 
surface. 
Examination of the efficiency of the Pasteur filter has been 
carefully done by a number of investigators, with results 
that may be summarized as follows: Water can be com- 
pletely sterilized by filtration through unglazed porcelain, 
but the filtration must not be continued for many days at a 216 
WATER-SUPPLY 
time. The length of time during which a sterile filtrate 
may be obtained will depend upon the temperature of the filter 
and its'contents. Thus, according to Freudenreich, at a tem- 
perature of 590 to 64° F. the filtrate was 
sterile for from fifteen to twenty-one days, 
at a temperature of 720 F. it was sterile dur- 
ing nine days, while at a temperature of 950 F. 
it remained sterile only five days.* 
Water-pressure is not a factor in causing 
germs to pass through the porcelain, for 
their method of penetration is one of develop- 
ment rather than a transporting of initial bac- 
teria; in other words, they "grow" through 
the filter. Even when the pressure is nil 
they accomplish the passage in the usual 
length of time. 
Horrocks denies this and says: " The B. 
typhosus is not able to grow through the 
walls of a Pasteur-Chamberland candle, and, 
if proper care be taken to prevent the direct passage of 
organisms through flaws in the material and imperfections in 
the fittings, the Pasteur-Chamberland filter ought to give 
complete protection from water-borne disease." f 
In this connection it must be noted that filtration will not 
stop the active principle of every disease, infantile paralysis 
for example. 
From a consideration of all the data given the line of man- 
agement for a " Pasteur " becomes plainly evident. The 
" candle " and its rubber packing should be removed at least 
once a week, thoroughly washed, and then boiled for half an 
hour before being reset in position. 
Especial care should be taken that there are no flaws in 
the " candle " and that the rubber packing make a tight fit, 
as otherwise the filtrate may pass around rather than through 
PASTEUR FILTER. 
* Centralblatt fiir Bakteriologie, xii. 240. 
f " An Introduction to the Bacteriological Examination of Water." By W. H. 
Horrocks. ARTIFICIAL PURIFICATION OF WATER 
217 
the porcelain. The filter should not be located in too warm a 
place. 
The lateral connecting the filter with the supply-main 
should be of tin or iron, as the slow action of the filter allows 
the water to remain for a considerable time in contact with the 
metal. 
Many types of household filters are of such character as to 
preclude the possibility of sterilization, and some of them it is 
even impossible to clean without the entire renewal of the 
filtering material. Such defects are necessarily fatal to proper 
filtration. The stone filters one sees at times, where the 
water is caused to drip through fine-grained sand-rock, or 
similar material, act as mere strainers and are absolutely 
unreliable. 
The author demonstrated that a sponge filter, after use for 
a month furnished a filtrate containing 121,200 bacteria per 
cubic centimetre, as compared with a count of 540 per cubic 
centimetre, in the unfiltered water. Household filters of many 
types are so seriously contaminated that a water cannot but 
be rendered worse by passage through them, and yet such 
appliances are in full use, and greatly trusted on account of 
the apparent clearness of the water drawn from them.* 
The common belief is that a filter, once established, is good 
for all time, and an amount of carelessness is permitted in other- 
wise well-organized establishments, that would almost stagger 
belief. 
For instance during the cholera epidemic at Lucknow in 
1894 out of 646 members of the East Lancashire Regiment 
143 were attacked by cholera and 92 died. Direct proof was 
had of the infection of the barrack-room filter, f 
Much stress is often laid upon the purifying effects of 
animal charcoal, and the great quantity of occluded oxygen the 
fresh charcoal contains fully justifies for a time the high praise 
* A very full report upon the various types of filters in use, by Woodhead and 
Wood, may be found in British Medical Journal, vol. ii, pp. 1053, m8, 1182, 
1375, and i486. 
f See "Applied Bacteriology," by Pearmain and Moor, p. 312. 218 
WATER-SUPPLY 
given it, but such material is nearly impossible to cleanse, and it 
has been repeatedly shown that a more objectionable appliance 
could scarcely be found, from a sanitary point of view, than a 
neglected charcoal filter. For instance, Frankland finds that 
in the case of such a filter having been in use a month, the 
filtrate contained 6958 bacteria per cubic centimetre, as against 
1281 per cubic centimetre in the unfiltered water.* 
A similar record is made by Percy Frankland, of water passed 
through a filter composed of six inches of animal charcoal, 
in a state of fine division: 
Period. 
Organism per Cubic Centimetre. 
Unfiltered Water. 
Filtered Water. 
Initial . . 
After 12 days 
After i month 
Very numerous 
2800 
1280 
none 
none 
7000 
For certain special uses, however, the carbon filter is val- 
uable. Thus the writer recommended passing an already 
filtered river-water through a layer of fine coke in order to 
remove the last trace of color which was not extracted by the 
use of alum. An absolutely colorless water was required for 
the manufacture of a particular grade of paper, and it was 
secured by such means. 
The good old household remedy of boiling a suspicious 
water is always available, but unfortunately it is but seldom 
applied. As an illustration of its value it is worth noting that 
during the cholera outbreak at Lucknow, referred to above, 
among the members of the East Lancashire Regiment, there 
was no cholera whatever in company E. This company was 
the only one which boiled the drinking-water.f 
Emergency purification methods have to be adopted upon 
occasion, as when accidental pollution of a distribution reser- 
* Chemical News, lii, 27. 
t Rideal, "Water Purification," p. 151. ARTIFICIAL PURIFICATION OF WATER 
219 
voir occurs, or when the force of circumstances demands the 
use of water known to be impure. This is frequently a mili- 
tary necessity. 
Disinfection of a small reservoir is best secured by wasting 
the water and sprinkling the sides and bottom with strong 
solution of bleaching powder. 
Such a treatment was extended to the street mains at Maid- 
stone, England, at the time of their great typhoid epidemic. 
Disinfection of the pipes was very thorough but so also was the 
damage to the pipe fittings, as the strength of the solution 
used was much too great. 
After the great flood of 1913 at Columbus, Ohio, Hoover 
and Scott disinfected the city mains with lime.* 
" All water pumped into the city was treated with 1 to 
2 grains per gallon of lime in excess of that necessary to neu- 
tralize the free and half-bound carbonic acid. This is harm- 
less, does not taste and there was no evidence, except by chem- 
ical test, to indicate any causticity in the water; Samples 
were collected from numerous points in the city, from which 
it was found that the caustic water had diffused throughout 
the system in a short time. The river water had 500,000 bac- 
teria per cubic centimetre and showed positive presumptive 
tests in one to fifty dilutions, while the samples from the mains 
were negative in 10 c.c. portions twenty-four hours after pump- 
ing was resumed. 
" When enough lime is added to water to absorb the free 
and half-bound carbonic acid and to precipitate the magnesium 
content, the bacteria of the colon and typhoid groups are 
killed in forty-eight hours after being so treated. 
" The germicidal action is effective in from five to twenty- 
four hours when an excess of | to 1 grain per gallon is added 
beyond that needed to reduce the temporary hardness to the 
lowest possible figure. 
"Intestinal organisms will not live in water containing no 
free or half-bound carbonic acid." 
The inability of B. coli or B. typhosus to live in the absence 
* See Eng'neer'ng Record, 67:395. 220 
WATER-SUPPLY 
of free or half-bound carbonic acid has been also shown by C. 
Haller.* 
Thresh sterilizes water for the army in the field by adding 
an excess of " bleach," at least one part of available chlorine 
per million parts of water, and, after an interval of fifteen min- 
utes, removing the excess of sodium thiosulphate.f 
Boiling the drinking water, with subsequent aeration to 
remove the flat taste, is often practicable during active service 
and it is efficient. 
* Chem. Zentr., 1914: i. 193; also Chem. Abstracts, viii. 2206. 
f Lancet, 1914: 809. CHAPTER IV 
NATURAL PURIFICATION OF WATER 
Nature disposes in sundry ways of the various elements 
of impurity added to water, but by far the most important 
of these is the process termed " nitrification." This is a 
change best established by infiltration through soil, a few feet 
of such passage being capable of doing more to restore a water 
to its original purity than miles of flow in an open channel. 
Nitrification is accomplished by bacteria whose function is 
to tear asunder the objectionable nitrogenous organic materials, 
secure their union with the oxygen of the air and thus convert 
them into harmless inorganic forms, which are at the same 
time valuable as plant-food. 
The conditions under which these little organisms can 
thrive must be met, otherwise the oxidizing action will quickly 
diminish or even entirely cease. Darkness is favorable and 
strong light is fatal to the process. A supply of free oxygen 
must be at hand, but, as has been pointed out by Drown, a 
small fraction of the amount normally present in the atmos- 
phere will quite suffice for complete nitrification. The pres- 
ence of a base is necessary to fix the nitric acid formed. 
The action of the nitrifying bacilli is mainly confined to the 
upper layers of the soil, i.e., to those portions subject to culti- 
vation, and it rarely occurs below a depth of six feet. This 
diminution in number with depth in soil occurs in the case of 
other bacteria as well. 
One feature of special interest is that the nitrifying organism 
does not thrive in presence of a great excess of organic matter. 
It cannot be successfully cultivated in either bouillon or strong 
sewage. Furthermore, it is noticed that where nitrification is 
once thoroughly established other germs tend to die out, prob- 
ably on account of lack of food-supply. 
221 222 
WATER-SUPPLY 
The great importance of this purifying process of nitrifica- 
tion will be better appreciated when we come to consider the 
question of ground-water, for it is at once apparent that our 
wells must receive large contributions from drainage material 
poured into and upon the soil; One thing must be ever borne 
in mind when depending upon the purifying action of soil, 
namely, that its power should not be overtaxed by excessive 
doses of sewage material, and that its use as a filter must always 
be permitted to be intermittent, so that a proper supply of 
oxygen may be provided. The importance of aerating the 
filtering soil between the successive applications of sewage has 
been abundantly shown by the Massachusetts Board of Health, 
and the advantages of so doing are demonstrated on the large 
scale upon all sewage farms. 
Sewage for the purpose of broad irrigation is conducted 
throughout the irrigated district in open conduits of earth, 
and small side-gates at intervals admit it to the furrows between 
the rows of growing plants; the gates being opened and the 
furrows filled whenever, in the judgment of the attendant, 
the vegetation can appropriate the dose. 
The face of the land is all divided into small sections, in 
places less than an acre in area, and each such division is flooded 
independently; an arrangement whereby the filtration of the 
sewage through the soil can be maintained entirely inter- 
mittent in character, and nitrification given abundant oppor- 
tunity for full development. 
Any surface-clogging of. the ground is avoided by suitable 
use of the spade. So far as the quantity and quality of the 
crops raised are concerned, they appear to be very near per- 
fection. 
The conclusions reached by the Massachusetts Board of 
Health are as follows: 
" The purification of sewage by intermittent filtration 
depends upon oxygen and time; all other conditions are second- 
ary. Temperature has only a minor influence; the organisms 
necessary for purification are sure to establish themselves 
in a filter before it has been long in use. Imperfect purifica- NATURAL PURIFICATION OF WATER 
223 
tion for any considerable period can invariably be traced either 
to a lack of oxygen in the pores of the filter, or to the sewage 
passing so quickly through that there is not sufficient time 
for the oxidation processes to take place. Any treatment 
which keeps all particles of sewage distributed over the surface 
of sand-particles, in contact with an excess of air for a sufficient 
time, is sure to give a well-oxidized effluent, and the power of 
any material to purify sewage depends almost entirely upon its 
ability to hold the sewage in contact with air. It must hold 
both sewage and air in sufficient amounts. Both of these 
qualities depend upon the physical characteristics of the mate- 
rial. The ability of a sand to purify sewage, and also the 
treatment required for the best results, bear a very close rela- 
tion to its mechanical composition." 
We have thus seen that nature makes abundant provision 
for the removal of pollution from water that is poured upon the 
soil by oxidizing the objectionable material indirectly through 
bacterial agencies. 
As to the efficiency of those means, so highly thought 
of by the people at large, and supplied wherever the water 
passes over rapids and falls, namely, agitation and aeration, 
not so good an account can be given. Does direct oxidation 
take place, and, if so, to what extent? With a view of obtain- 
ing light upon this question, extended series of investigations 
have been undertaken with results which have produced a nega- 
tive reply. Perhaps the best known are Prof. Leeds' exam- 
inations of the water of the Niagara River before and after 
passing Niagara Falls. He found no improvement resulting 
from the passage over the cataract, and it is but reasonable 
to conclude that direct oxidation does not serve as a factor of 
any considerable importance in the purification of polluted water. 
In this connection, and as a result of his own investiga- 
tions, Frankland says: 
" We are aware that ordinary oxygen does not exercise 
any rapid oxidizing power on organic matter. We know that 
to destroy organic matter the most powerful oxidizing agents 
are required. We must boil it with nitric and chloric acid and 224 
WATER-SUPPLY 
the most perfect chemical reagents. To think to get rid of 
organic matter by exposure to the air for a short time is 
absurd." 
Bearing in mind the known powers of resistance of the 
various bacteria, it is difficult to conceive of any appreciable 
diminution in their numbers resulting from a short-time exposure 
of the water in the form of spray. 
Neither is it easy to see that the beneficial labors of the 
nitrifying bacillus can be materially aided by the momentary 
passage of a fall, when we remember the small percentage of 
dissolved oxygen required for the fulfilment of its task. 
That the said nitrifying bacillus can, under any circum- 
stances, accomplish in a water the quantity of work expected 
of it in a soil is, of course, not to be hoped for. 
It must not be assumed, however, that the old and firmly 
planted belief of the people is entirely false, and that aeration 
is without any value whatever. Keeping a water well satu- 
rated with atmospheric oxygen, either by spraying it in form 
of a fountain, as at Rochester and elsewhere, or by pumping 
air into it, unquestionably renders less likely the growth of 
some forms of algae, with their accompanying odors and 
tastes, and also removes, by direct displacement, any foul 
gases or volatile products already in solution. 
The Kensico aerator of the new supply for New York city 
consists of a concrete basin 460 by 240 feet in area from the 
floor of which the water is thrown into spray through 1750 
nozzles. 
It is undoubted wisdom to encourage the existing tendency 
to aerate public waters, but the true action of such aeration 
must be always kept in mind, to the exclusion of false and 
exaggerated notions of its value. 
Sedimentation is another purifying process upon which wide 
dependence is placed. Its consideration properly comes under 
a discussion of lake and reservoir waters, but a word should 
be said here with reference to what may be expected of it in 
the cases of streams and rivers. NATURAL PURIFICATION OF WATER 
225 
With a view of determining to what extent sedimentation 
can be depended upon for the purification of streams, the fol- 
lowing inquiry was undertaken. 
Upon four different occasions (covering various conditions 
of medium, high water, and flood) samples were analyzed from 
that section of the Hudson River extending between Troy and 
Albany. The stations at which samples were taken are sit- 
uated over one mile apart, beginning at State Street, Troy, 
and ending at Albany, five miles below. Two samples were 
taken at each station during ebb-tide and in mid-channel; 
one two feet from the surface and the other, as near as could 
be judged, two feet from the bottom. 
The examination of the samples showed sedimentation at 
all stages of the river, the average being nearly constant through- 
out the entire distance. 
Such sedimentation was, however, found to be decidedly 
small. An idea of the amount deposited may be obtained from 
the fact that average differences for total solids between the 
upper and the lower samples at the end of the second mile was 
3.47 per cent of the total solids in the upper sample. 
A review of the evidence furnished by the inquiry leads to 
the belief that sedimentation as a source of river purification in 
streams such as the Hudson is not nearly so valuable as has 
been heretofore held. And it would appear, moreover, that a 
river mainly clears itself by "running out " the suspended silt 
rather than by depositing the same upon its bed. 
So far as the removal of bacteria from river-water by sedi- 
mentation is concerned, it must be remembered that, their 
specific gravities being only slightly greater than unity, they 
sink but slowly in quiet water, and of necessity still less rapidly 
in that which is moving. That specific germs do not all sub- 
side during even long distances of flow may be inferred from the 
typhoid statistics already given. 
Freezing water does not completely purify it, as was formerly 
supposed; neither is it correct to assert that bacteria in general 
resist low temperature, as was more recently held to be true. 226 
WATER-SUPPLY 
Under the chapter on " ice " it will be pointed out that 
mechanical exclusion and lapse of time are large factors tending 
toward purification during the formation and storage of ice; 
We now recognize that great mortality occurs among bacteria, 
because of low temperature and pressure, during the process of 
freezing; and we further find that such mortality will vary 
with the type of the organism. 
As a single instance: ordinary tap water lost sixty per cent 
of its total bacteria by simply freezing solid. The whole 
volume of the water was turned to ice; the ice was then imme- 
diately melted and the water examined for total bacteria, 
with the result as above stated. 
Mechanical exclusion and lapse of time did not enter in 
this case and the resulting mortality was due to low tempera- 
ture and pressure. 
This experiment simply shows that a high mortality follows 
freezing, but the actual percentage is of no great significance, 
as a variety of numerical results can be obtained, some of them 
showing a mortality reaching to nearly 100 per cent. 
The inadequacy of ice formation as a protection against 
bacterial pollution is more fully referred to upon another 
page. 
It remains to say a word concerning the purifying action of 
sunlight supplementary to what has been already given on 
page 65. 
Very exhaustive investigations by Prof. H. Marshall Ward 
show light to have great germicidal action and that 11 the rays 
which kill the bacteria are the blue and violet ones. The 
infra-red, red, orange, yellow, and green are without effect." 
" This explains why these organisms are destroyed so much 
more rapidly by the light of the summer sun than in winter; 
why a clear blue sky is so much more effective than a hazy 
one, and why direct sunlight acts so much more quickly than 
reflected or diffused daylight." * 
Investigations upon this topic are as yet uncompleted, 
* Chemical News, Ixx. 243. NATURAL PURIFICATION OF WATER 
227 
but enough has been done to show the marked toxic effect of 
sunlight upon bacterial life and its consequent aid to the effort 
of the sanitarian. For full and detailed information upon the 
subject the reader is referred to the work of Percy Frankland.* 
Stated roughly, sunlight is fatal to bacteria sooner or later, the 
intensity of the action depending upon the kind of germ and 
the brightness of the light. 
Buchner gives a graphic illustration of the action of light, 
using the typhoid bacillus for the demonstration,! the value 
of such experiments is far-reaching, and suggestions of sanitary 
importance naturally follow. 
The bearing this point has upon the influence of sunlight 
upon the self-purification of streams is at once apparent; but 
it must not be forgotton that a comparatively thin layer of 
water will cut off an immense deal of the germicidal power of 
sunlight, and we must consequently restrain our tendency to 
exaggerate the beneficial action. 
In this connection it is worth while studying the antiseptic 
action (recorded by Procacci) of midday sunlight, in June, 
upon bacterial life contained in drain-water forty inches deep. 
The light was passed through the water vertically, side-light 
being excluded, and the time of exposure was three hours. 
Comparison tests, kept in darkness, were also made. The 
results were as follows per cubic centimetre: 
Before Exposure. 
Sunshine. 
Darkness. 
Surface 
2100 
9 
3103 
Centre 
2103 
10 
3021 
Bottom 
2140 
2115 
3463 
The sterilizing action of light upon the upper portion of the 
water is thus seen to have been marked. 
After all the bactericidal action of sunlight has been 
acknowledged we must recognize that, as a sanitary proposi- 
tion, the beneficial results to be obtained from it as a water 
purifier are small indeed. Its power of penetration is too low, 
* "Micro-organisms in Water." 
f "Einfluss des Lichtes auf Bakterien." Centralblatt f. Bakteriologie, xi. 781. 228 
WATER-SUPPLY 
especially in turbid or in colored waters or when winter has 
covered the rivers with layers of ice. 
Even the ultra-violet ray is efficient only in clear and nearly 
colorless waters, and then to only very moderate depths. 
Self-purification of Streams 
Pettenkofer expressed the opinion * that 11 ordinary sewage 
may be, without hesitation, turned into any river or brook 
whose volume is fifteen times the volume of the sewage, and 
whose velocity is not less than that of the stream of sewage. 
Under these circumstances the necessary dilution and self- 
purification take place after a short flow." 
If this were only true, the vexed question of sewage dis- 
posal would be largely disposed of, and enormous sums of 
money now expended in such disposal would be saved. That 
it is very far from being safe practice is evidenced by such 
statistics as have already been quoted showing the serious 
pollution of large rivers by small streams of sewage inflow. 
It has been shown (page 28) that twenty-six miles of flow 
was not enough to protect Albany from the contaminated sew- 
age of Schenectady, even when the rivers in question were so 
large as the Mohawk and Hudson, and with the high " Co- 
hoes " falls on the route. 
Prof. Sedgwick gives an instance of similar carriage by the 
Merrimac River: 
TYPHOID INFECTION CARRIED TWENTY-FIVE MILES BY RIVER 
Cases Reported. 
Deaths. 
Lowell. 
Lawrence. 
Newb'pt. 
Lowell. 
Lawrence. 
Newb'pt. 
November, 1892 
19 
14 
O 
3 
4 
0 
December, 1892 
70 
32 
4 
IO 
9 
I 
January, 1893 
38 
72 
28 
IO 
3 
3 
February, 1893 
14 
23 
9 
7 
12 
0 
March, 1893 
4 
4 
I 
-From Mass. Reports, 1892. 
* Fischer, "Das Wasser," 268. NATURAL PURIFICATION OF WATER 
229 
11 In the eight months preceding August, 1892, two cases 
of typhoid were reported in Newburyport, in the subsequent 
five there were ten; twenty-eight cases in January, 1893, were 
thus very unusual. These cases appeared in the same month, 
but earlier than the increase in Lawrence; they were therefore 
due to infection from Lowell, more than twenty-five miles dis- 
tant. The people had warning of the danger from Lawrence." 
Professor Williams told the writer that in the summer of 
1892 a serious epidemic occurred in Detroit, so that the typhoid 
death-rate in that year was exceeded by that of only four 
American cities. The cause was finally traced to Port Huron, 
sixty miles north. The sewers of Port Huron discharged into 
Black River, a sluggish stream with very little flow. In the 
spring of 1892 work was begun by the government upon the 
dredging of Black River to improve it for navigation, and 
deposits of foul material from the years' accumulation of 
sewage in the stagnant reaches were removed by dredging, 
some of these deposits being eight to twelve feet in depth. 
This material was taken away in scows and dumped into the 
swift current of the St. Clair River. Allowing for the time 
necessary for the water thus polluted to reach Detroit and for 
the period of incubation of the typhoid germ, the course of the 
fever in Detroit exactly coincided with the dredging of the pol- 
luted material and its deposit in the river, sixty miles above. 
In their sixth report (page 138) the Rivers Pollution Com- 
missioners of Great Britain said: " We are led to the inevitable 
conclusion that the oxidation of the organic matter in sewage 
proceeds with extreme slowness even when the sewage is mixed 
with a large volume of unpolluted water, and that it is impos- 
sible to say how far such water must flow before the sewage 
matter becomes thoroughly oxidized." 
The inference contained in this old report is not entirely in 
accord with modern experience. We no longer look entirely 
to the chemical examination of our information, and we recog- 
nize other elements of harmfulness than merely dead organic 
waste material, and other means of oxidation than direct atmos- 
pheric action; but we believe, as they did, that self-purification 230 
WATER-SUPPLY 
of streams is a process not to be implicitly relied on, and that 
simple dilution enters largely into the safety factor of those 
who drink water from a polluted river. 
When the question of the self-purification of streams first 
came to the fore, Pettenkofer's theory held undoubted sway, 
and short distances of river-flow were considered entirely 
adequate to the removal of gross pollution. Later on the 
development of bacteriology threw discredit upon this view, 
and it was -held that a water once seriously polluted could never 
again be safely turned to domestic use unless artificially puri- 
fied. 
The investigations connected with " The Chicago Drainage 
Canal Case " in the Supreme Court of the United States, dealing 
with the alleged pollution of the Illinois and Mississippi Rivers 
by the opening of the Chicago drainage-canal have tended to 
again modify our views, and have caused us to admit that 
practical stream-purification is a fact provided the length of 
flow be sufficiently great. The distance from Chicago to St. 
Louis by water is 357 miles, and concerning the situation at 
the latter place the opinion of the court was that Chicago 
was not at fault so far as the production of typhoid fever in 
St. Louis was concerned. In other words, the Illinois and 
Mississippi Rivers sufficiently dispose of Chicago's sewage by 
natural means during the above-mentioned length of run. 
The record in that case was exceedingly lengthy and.even 
the abstract and summation of it made a large volume; some 
few points, however, may not be out of place here. 
With reference to the widely quoted 11 Bacillus Prodigiosus 
experiment," wherein 107 forty-gallon barrels of a culture 
of the bacillus containing from 500 to 1000 million bacilli per 
cubic centimetre were emptied into the Chicago Drainage Canal, 
and the organism afterwards sought for at the St. Louis intake; 
the opinion given was, that in view of the enormous number 
of the bacilli introduced and the small number (5) found, the 
result of the experiment show a very considerable death-rate 
among the germs employed in the test and that, by inference, 
at least as large a death-rate of the bacilli of typhoid fever, NATURAL PURIFICATION OF WATER 
231 
which they represented, would occur if the latter were exposed 
to the same conditions. 
No little time was devoted to a consideration of the Ion- 
ILLINOIS RIVER 
AND ITS 
PRINCIPAL TRIBUTARIES 
gevity of the typhoid bacillus; contrasting its power of retain- 
ing life for months while in the human body, or while in moist 
soil, with the shorter duration of vitality when immersed in 
water; the latter period being measured by days. 
Recognition was naturally had of the influence of the 232 
WATER-SUPPLY 
relative purity of the water under consideration upon the 
result and the tendency of the great density of germ life in 
polluted waters to inhabit a continuance of the typhoid organism. 
The view was held that in a relatively pure water its chance 
for life is greater than in one where it has to fight for exist- 
ence amid saprophytic bacteria, and the question of the lon- 
gevity of the germ in water was met by the statement that it 
dies out therein more or less quickly en masse, although some 
individual cells, being more resistant, may remain alive for a 
longer period. 
When contrasting the different results obtained while 
operating upon the experimental scale, using the glass contain- 
ers common to the laboratory, and comparing them with those 
secured under conditions more nearly approaching what are 
found in nature, reference was made to the experiments con- 
ducted by the Mass. Board of Health (1902, page 259) touch- 
ing upon the longevity of the Bacillus Coli communis. The 
vessel employed by the Board of Health was a large tank twelve 
feet deep. Under those circumstances the life of the bacillus 
in Merrimac river-water was found limited to eleven days. 
By inference the typhoid bacillus would not have lasted so 
long. In other words, as we approach the conditions found 
in nature the opportunities for the destruction of the patho- 
genic organism seem to be greater than in those obtaining dur- 
ing during a laboratory experiment. 
Dilution as a means of purification was considered and the 
claim was advanced that although no destruction of organisms 
can be expected, yet the risk incurred by the man who con- 
sumes a pint of water at a draught is certainly diminished through 
dilution by reason of the fixed number of germs being distributed 
throughout an increased bulk of water. To the extent therefore 
that " purification " can be interpreted as a lessening of the 
degree of danger, to that extent dilution is equivalent to partial 
purification. 
A point upon which stress was laid was the possibility of 
typhoid organisms accumulating in the mud of rivers by sedi- 
menting from up-stream sources, after a sort of " placer " 233 
NATURAL PURIFICATION OF WATER 
fashion, and later becoming " scoured " out during times of 
flood to the damage of the populations below. As typhoid 
bacilli will live a long time in moist soil it was asked if a similar 
vitality were not to be expected in the sludge of river bottoms? 
There are British statements made which answer this inquiry 
in the affirmative, but they do not appear to be supported by 
experimental evidence. On the other hand, the Massachusetts 
Board of Health experiment already quoted showed the absence 
of the Colon bacillus from the sediment collected at the bottom 
of the tank, the examination having been made after an interval 
of fifteen days. 
Should accumulations of virulent germs collect upon river 
bottoms from up-stream sources and then later be washed out by 
floods,widespread disease would follow throughout the sec- 
tions below, yet upon inquiry we do not find recorded as close 
a relationship between floods and epidemic typhoid as we should 
reasonably expect under such assumption. The Detroit typhoid 
following dredging at Port Huron, already referred to on page 
229), would seem evidence in support of the "placer " theory- 
but if that theory were founded upon a general truth there would 
be many more like instances of down-stream disease resulting 
from the scouring of stream-bottom deposits. It is easy to 
understand an increase of typhoid following the rise of a stream 
like the Tees in northern England, but in that case the accumu- 
lations of filth which were swept seaward by the swollen river 
were not the result of sedimentation, but consisted of night 
soil and other matters that had gathered upon the dry and rocky 
foreshores during periods of drought. Such a case is scarcely 
comparable with the gradual accumulation of river mud, for 
we know that typhoid germs will live for long periods of time 
in fecal deposits. In soil the water exists as " moisture " 
only, while in river sludge it is present in quantity, and oppor- 
tunity is thereby given for the solution and adverse action of 
the toxins of antagonistic organisms. This view was, the author 
believes, first advanced by Jordan, who backed it by experimental 
support, yet in stating it the fact should not be forgotten that 
Sedgwick in 1893 succeeded in growing typhoid bacilli in filtered 234 
WATER-SUPPLY 
sewage which contained no germ life of any kind but was pre- 
sumably rich in saprophytic toxins. 
In an effort to show the connection, if any, between the 
prevalence of typhoid fever in the two cities of Chicago and 
St. Louis, the following charts were constructed exhibiting 
JANUARY 
FEBRUARY 
MARCH 
APRIL 
MAY 
JUNE 
JULY 
AUGUST 
SEPTEMBER 
OCTOBER 
NOVEMBER 
DECEMBER 
DEATHS FROM TYPHOID FEVER AS REPORTED IN CHICAGO AND ST. LOUIS OFFICIAL 
HEALTH REPORTS. DEATHS IN ST. LOUIS ARE PLOTTED ONE MONTH IN 
ADVANCE, I.E., FEBRUARY DEATHS ARE PLOTTED AS OF JANUARY. (THE 
DRAINAGE CANAL WAS OPENED IN JANUARY, IQOI.) 
the deaths from the disease by months. The deaths in St. 
Louis are plotted one month in advance; thus the February 
numbers for St. Louis are made to correspond with the January 
readings for Chicago in order to allow for the time consumed 
by the bacillus in making the journey by water between the 
two cities, and for the further time required to develop a fatal 
result after the bacillus had been swallowed. To the author's NATURAL PURIFICATION OF WATER 
235 
mind there is a lack of parallelism between the curves as shown, 
nor is the case bettered by increasing the time allowance above 
noted to beyond one month, all of which does not point toward 
Chicago typhoid having been causative of that found in St. 
Louis. 
The writer sought to lay stress upon the importance of the 
element of " time " in the self-purification of water, and to call 
attention to the greater probability of St. Louis importing 
its typhoid from points of pollution nearer at hand than Chicago* 
In support of this view data were offered which were later em- 
bodied in a paper before the New England Water Works Asso- 
ciation entitled " Relation of Intensity of Typhoid Fever to 
Character of Water Carriage " (Vol. 19, page. 412). Space 
does not here permit of extended reference to that paper. Suffice 
it to say that in a mixed epidemic which the writer had oppor- 
tunity to study (i.e., an epidemic caused by infected water 
from two sources) it was found that 27.8 per cent of the typhoid 
cases were severe in type where the disease was traced to a 
distant origin, while 56.5 per cent was severe when the fever 
was caused by water from a local source. It would appear from 
this that support is given to the general proposition that the 
danger of illness and death will vary inversely as the length of 
time elapsing between the entrance of polluting material and 
the drinking of the water. Or, stated in other words, the 
typhoid bacillus, under prolonged struggle for existence, will 
lose some of the vigor of its life and will become incapable of 
producing its normal amount of poison. Under such circum- 
stances either no intoxication results or else a mild type of dis- 
ease develops among those expecially susceptible to invasion. 
However much we are inclined to accept the Chicago 
Drainage Canal data as an addition to our knowledge concern- 
ing the changes induced by river-flow, one cannot but feel 
that no upsetting blow has been given to the general proposi- 
tion that specific infection may cover long distances by water- 
carriage, and that it would be very rash to prophesy for any 
particular case just what the limits of such carriage would 
be until after thorough investigation had been made. 236 
WATER-SUPPLY 
As dealing alone with the chemical changes occurring in 
polluted water during stream-flow, a former investigation 
conducted by Professor Long is well worthy of attention. He 
instituted a series of analyses of the dilute sewage contained in 
the Illinois and Michigan Canal. It is to be noted that this canal 
received its supply of water (or rather dilute sewage) at Bridge- 
port, where the pumps delivered to it the filthy water of the 
Chicago River, contaminated with a great portion of the sewage 
of Chicago. From Bridgeport the water " flows along the 
level to Lockport, twenty-nine miles below, requiring about a 
day for its passage." It received no dilution on the way and 
was frequently agitated by passing boats. " After passing 
Lockport the water descends to Joliet through four locks, and 
falls over a dam seven feet in height to a point of collec- 
tion. There is a fall of 58.2 feet in a distance of four miles, 
and no dilution takes place on the way." The examina- 
tions of this canal-water were both numerous and thorough, 
and judging from the mean results there is good ground for 
the statement that very considerable self-purification of a 
chemical character took place during the flow of thirty-three 
miles. 
The writer has long been of the opinion that what may be 
true for dilute sewage does not hold good as we approach the 
limit of potable water. 
In other words, so far as the improving of a water by the 
natural processes of oxidation is concerned, the rate of such 
improvement varies directly as the amount of sewage contamina- 
tion. Given a stream with a certain amount of pollution, the 
per cent of such pollution which disappears per mile of flow 
will continually decrease as the stream flows on. 
Plotting Dr. Long's figures in graphic form they assume the 
shape shown on the following charts. The change in lake- 
level at various dates, together with other disturbing influ- 
ences, caused comparatively clean water to reach the pumps 
at times, and we therefore are furnished with data governing 
the purification of several variously contaminated waters while 
flowing under constant conditions. NATURAL PURIFICATION OF WATER 
237 
FREE AMMONIA. 
BRIDGEPORT TO LOCKPORT. 
SELF-PURIFICATION IN ILLINOIS AND MICHIGAN CANAL. 
ALBUMINOID AMMONIA. 
BRIDGEPORT TO LOCKPORT. 
SELF-PURIFICATION IN ILLINOIS AND MICHIGAN CANAL. 238 
WATER-SUPPLY 
It will be noticed that the rate of purification per mile for 
the more grossly contaminated samples is much greater than 
that for those comparatively pure. 
Even the best of these waters of the Illinois and Michigan 
Canal was very far from being potable, and we may conse- 
quently look for still further reduction in the purification rate 
as we near the potable limit. 
We have seen that the amount of oxygen dissolved in a 
REQUIRED OXYGEN. 
BRIDGEPORT TO LOCKPORT. 
SELF-PURIFICATION IN ILLINOIS AND MICHIGAN CANAL. 
water need not be large in order to permit the purification 
changes inaugurated by nature to go on; but in the event of 
the supply of oxygen being entirely cut off, putrefactive reac- 
tions are set up with very undesirable results. 
A case of this kind is reported by Dr. Leeds as occurring 
during the winter of 1882-3. The exceedingly bad taste and 
smell of the Philadelphia water was found by him to have been 
due to a superabundance of putrescible material, at a time NATURAL PURIFICATION OF WATER 
239 
when the dissolved oxygen was unusually small in quantity. 
The rainfall of the late autumn and early winter had been 
very slight, thus producing a low state of the river. Polluted 
water in extraordinary quantity had been admitted by the 
emptying of sundry dams and canal levels, and the atmos- 
phere was cut off from direct action upon the river by a 
continuous coating of ice. Under these circumstances the 
foul-smelling compounds well known to form when organic 
matter decomposes out of contact with air were produced 
in large quantity, to the great discomfort of the water- 
consumers. So considerable in amount were the gaseous 
products in this instance that it was possible to ignite them 
as they escaped from holes in the ice. The flame pro- 
duced by igniting the gas issuing from a penknife puncture 
of the white hollows under the ice is described as being 
usually six inches high, but once it was 11 fully a yard 
high." * 
A curious instance of similar character was reported in the 
Chicago papers of November 2, 1894, Refuse matter had 
accumulated in so great quantities in the Chicago River that 
the available oxygen was far too small in quantity for its 
proper oxidation. Gaseous and inflammable products of sub- 
aqueous putrefaction resulted, which upon ignition at the 
surface became almost dangerous to shipping. " The tug 
A. Mosher was towing the schooner Ford River out of the 
South Fork, when both boats were surrounded by the fire, 
which was consuming the gases rising to the surface in huge 
bubbles." 
Judged from the bacteriological standpoint, a considerable 
monthly and seasonal variation will be observed in the purity 
of running streams. During the low water of summer, although 
the relative volume of sewage-inflow is large, yet the absence 
of surface-washings, due to storms or melting snow, usually 
causes diminution in the total number of bacteria present; 
* J. Frank. Inst., Ixxxvi. 26. 240 
WATER-SUPPLY 
for instance, Percy Frankland gives the following counts of 
bacteria per cubic centimetre in Thames water collected at 
Hampton: 
January 92,000 
February 40,000 
March 66,000 
April 13,000 
May 1,900 
June 3,500 
July 1,070 
August 3,000 
September.. ..'. 1,740 
October 1,130 
November 11,700 
December 10,600 
Jordan was the first to point out that improvement of a 
chemical type and of a bacterial character, during the flow of a 
stream, do not follow parallel courses. 
Objectionable bacteria may die out during a run shorter 
than that required to oxidize non-living organic matter. 
The seasonal variation in total count of bacteria is also 
partly due to the greater number of enemies the bacteria have 
to face during the warmer months. Both plants and animals 
which limit their growth are plenty in summer and relatively 
few in winter, when the water is covered by ice. There are 
many " bacteria-eaters," such as the large family of 11 Rotifers " 
or " Wheel animals," which in turn are eaten by animals of 
larger growth. 
Dr. Ruediger, Director of the State Public Health Labora- 
tory, University of North Dakota, Grand Forks, N. Dak., 
prepared a table showing variation of number of B. coli per 
centimetre in the water of Red Lake River with the change of 
season; the figures are averages of four analyses of samples 
collected at intervals of six hours: NATURAL PURIFICATION OF WATER 
241 
Date. 
B. Coli, per c.c. 
Volume of Water 
in River: Cu.ft. 
per Sec. 
Calculated B. 
coli per c.c. if 
Volume of Water 
Always = 750 
Cu.ft. 
River 
Not Covered 
with Ice. 
June 18, 1909 
21 
1120 
3-0 
July 12, 1909 
31 
782 
3-5 
Sept. 11, 1909 
61 
1000 
7-3 
June 4, 1910 
2 
1180 
3-3 
June 28, 1910 
51 
75° 
5-5 
Aug. 15, 1910 
6 
300 
2.0 
River Covered with 
Ice (average 
thickness 30 
ins.) 
Jan. 18,1910 
151 
77o 
15-5 
Mar. 1, 1910 
173 
610 
14.0 
A long, full article will be found in Engineering News, 
August 31, 1911 on "The Significance of Flora and Fauna in 
Maintaining the Purity of Natural Waters, and How they 
are affected by Domestic Sewage and Industrial Wastes." * 
Translated from the German by Emil Kuichling. 
At times it chances that some substance of especially marked 
smell or taste becomes mingled with the general mass of sewage 
contaminating a water, and a lively sense of pollution imme- 
diately takes possession of the consumers. For instance, 
some years since, a paper-mill dumped refuse containing a 
little carbolic acid, into a tributary of the Passaic River, at 
a point above the common intake of the cities of Newark 
and Jersey City. So strong was the odor and taste communi- 
cated to the supplies of the two cities that use of the water 
for drinking purposes was for a time discontinued. The 
quantity of carbolic acid consumed by each individual using the 
water was well-nigh infinitesimal and beyond possibility of 
doing harm, yet serious objection was made to the 11 pollu- 
tion " of a stream which was already laden with the sewage of 
fifty thousand persons' residing at the city of Paterson, only 
eighteen miles above. 
Again: The city of Cleveland, 0., takes its Lake Erie water 
* See also Whipple's "Microscopy of Water" for influence of higher fish 
life. 242 
WATER-SUPPLY 
from an intake situated beyond the breakwater, and the city 
sewage passes into the artificial harbor and thence is delivered 
into the lake at a point well down the east shore. It might 
be thought that there was little chance of sewage working 
so far westward agains(t the general trend of current, but it 
so happens that oily material from the Standard Oil Company's 
works constitutes a portion of the city sewage, and it has been 
noticed that a petroleum taste is given the water when the 
direction of the wind causes an accumulation of sewage in the 
harbor, which sewage is afterwards permitted to rapidly escape 
into the lake upon change in the weather. In each of these cases 
the special material was but a harmless indicator of the pres- 
ence of the unrecognized but far more dangerous sewage pol- 
lution, yet the public strongly objected to the one and calmly 
accepted the constant presence of the other. 
It will be remembered that some years ago, about 1879, 
a portion of Boston's water supply was considered in great 
danger of contamination with sulphuric acid, owing to the 
burning of chemical works situated upon one of the tributaries 
to Mystic Pond, whereby some fifty tons of oil of vitrol were 
washed into the stream. Mystic Pond is about eight miles 
below the site of the works, and mill-ponds intervene. Marked 
acidity was noticed at varying points in the course of the 
stream and intervening ponds, but no trace of acid remained 
by the time Mystic Pond itself was reached. 
This instance has been often dwelt upon to show how 
nature may take care of even the stable inorganic additions 
to surface-waters. CHAPTER V 
RAIN, ICE AND SNOW 
Saussure held that a part of the water raised into the 
atmosphere resembles soap-bubbles. 
Clouds he considered to be composed of small vesicles, of 
which water formed the envelope. And he further believed 
that every vesicle that rises from the sea must contain a small 
quantity of the solid matter which was dissolved in the sea- 
water; the proportion of solid matter taken up in a given space 
varying according to the relative proportions of these original 
vesicles that enter into the composition of the clouds.* 
The more modern view is stated by Douglas Archibald: f 
" Rain is the final stage of condensation of vapor back into 
water, of which cloud is a half-way stage. The mist which 
composes a cloud is formed of tiny drops of water about 
inch in diameter. It used to be a puzzle to explain how these 
water-particles were sustained, and it was at one time supposed 
that cloud-particles were hollow. We know now that this is 
neither necessary nor true, since very small particles, even of 
gold, will remain suspended for a long time in air; the finer 
the particles the longer they take to fall. A slight upward 
motion of the air is therefore enough to keep them balanced. 
As condensation proceeds these particles grow larger by fresh 
coatings of water, and the larger ones fall down against the 
smaller and mingle with them until large drops from to TV 
inch thick form, which are no longer capable of being sus- 
pended and fall to the earth. Snow forms when the tempera- 
ture at which this further stage of condensation occurs is 
below freezing-point. 
* Angus Smith, "Air and Rain," p. 233. 
f "The Story of the Atmosphere." 
243 244 
WATER-SUPPLY 
" Hail is frozen water-drops. It is believed that the rain- 
drops formed in one part of a storm are carried upward by 
powerful ascending currents (twenty-five miles an hour is 
enough to sustain large drops) into higher regions of the atmos- 
phere where they are solidified by the excessive cold, and 
being carried over with the overflow which takes place near the 
top, fall down until they are redrawn into the interior of the 
storm and again whirled up aloft. Receiving alternate melt- 
ings and freezings, and growing larger with each circuit they 
make in the atmospheric churn, they are finally thrown out on 
either side of the storm-centre. This explains the fact that 
in a traveling hailstorm there are two bands where hail falls 
on either side, while under the center it is often found that 
only rain has fallen." 
Hailstones of great size occasionally occur; thus on June 
23, 1906, a violent hail-storm swept over Staten Island, N. Y., 
during which hailstones fell measuring nine inches in circum- 
ference.* 
The presence of solid matter in rain-water is accounted for 
by the large quantities of dust of all kinds continually carried 
into the atmosphere by the winds, and washed therefrom by 
the falling drops. In the vicinity of large towns various prod- 
ucts of incomplete combustion and of industrial waste are 
added to the atmospheric impurities, and are precipitated along 
with the more commonly occurring dust; f while in the vicinity 
of the sea the amount of common salt present increases mate- 
rially. 
The presence of soot in the air causes increase in the rain- 
water of such impurities as sulphuric acid and ammonia, by 
what appears to be direct absorption of such material by the 
soot. This is shown by Mabery in the following analysis of 
the air of Cleveland, Ohio: J 
* Science, July 23, 1906. 
f The English Alkali Act permits the presence of HC1 to the limit of 0.2 
grains, and of S03 to the limit of 4 grains in each cubic foot of air, taken at the 
foot of the stack of such industrial plants as generate these waste products. 
$ J. Am. Chern. Soc., xvii. 3. RAIN, ICE AND SNOW 
245 
WEIGHT IN MILLIGRAMMES PER LITRE OF AIR 
Soot. Sulphuric Acid. Ammonia. 
87.5 15.2 .070 
45.2 6.3 .OIO 
III.3 21.2 .120 
41.8 I3.9 .003 
On September 8, 1894, there occurred the first rain, after 
the longest period of drought that had been experienced in the 
State of New York during forty years. Many forest fires had 
occurred and the atmosphere had been exceedingly hazy for 
weeks. 
The author collected rain-water, on the above date, in the 
Catskill Mountains, and the quantity of oily, sooty material it 
contained was very striking. 
The amount of this soot from the air may be very great, thus 
Kershaw found * as much as 25 tons per square mile per year 
in a country district in England and up to 539 tons per square 
mile per year in a manufacturing town. 
Baskerville and Weller give the analysis of a " black rain " 
colored by soot which fell at Louisburg, N. C., in March. 
19004 
A red rain, whose color was produced by dust of possibly 
a cosmic origin, fell upon a number of places in central and 
southern Europe between March 11 and 13, 1901. Other 
abnormal precipitations such as salt rain, rain colored by 
volcanic ash or by desert sand, and the " bloody snow " pro- 
duced by growths of Palmella sanguinea, have been mentioned 
by Phipson in his " Researches on the Earth's Atmosphere." | 
Averages for many years, showing amounts of ammonia and 
nitric nitrogen carried down by rain at Paris and in different 
parts of England, may be found in the Report of the Mont- 
souris Observatory for 1897. 
* Surveyor, 43:462. 
f Science xv, 1034. 
J See also Chemical News, Ixxxiii. 159. 246 
WATER-SUPPLY 
The various germs floating in the air, and ready to be 
carried down by the first shower, do not play a material role 
from the water-supply standpoint, partly because of the im- 
probability of their being pathogenic in character, and partly 
because of the germicidal power of the direct sunlight to which 
they have been so thoroughly exposed. 
Nevertheless it may be of passing interest to give the official 
figures issued by the Montsouris Observatory, Paris: 
BACTERIA PER CUBIC METRE OF AIR AT MONTSOURIS 
(Average for ten years) 
January 160 
February 145 
March 225 
April 310 
May 305 
June 355 
July 465 
August 455 
September 310 
October 190 
November 195 
December 165 
Mean 275 
A comparison of country and city air shows the following 
number of bacteria per cubic metre: 
Montsouris 275 
Center of Paris 6040 
There is a daily maximum of bacteria at 2 p.m. and a mini- 
mum at 2 a.m. 
It may be worth adding that Miquel gives the following 
comparison, in terms of bacteria per cubic metre, between the 
air of the Paris sewers and that of the public streets: RAIN, ICE AND SNOW 
247 
• ■ 
Air of the Sewers. 
Air of the Streets. 
Bacteria. 
Molds. 
Bacteria. 
Molds. 
Winter 
2385 
4050 
3,210 
599 
Snrin? 
7165 
2330 
11,085 
865 
Summer 
5110 
2730 
12,070 
2340 
Autumn 
5400 
1550 
7,365 
2320 
Mean 
5015 
2665 
8,435 
1530 
The monthly variation in the chlorine contained in rain- 
water collected at Troy, N. Y., is given in the following table, 
the determination having been made in a mixture of the entire 
rainfall for each month: 
January 2.50 per million 
February 1.07 
March 1.55 
April 0.75 11 
May 1.25 11 
June 1.15 
July 1.05 
August 2.00 " 
September 0.60 
October 3.00 
November 2.25 11 
December 2.50 
Mean 1.64 per million 
Even casual inspection will often show that rain-water is a 
long way from being pure, and,.high as this "water from the 
heavens " is rated in the public mind, it is frequently polluted 
to an extent quite surprising to the collector of the supply. 
The author has often noted the confidence with which 
people will make use of water from a foul cistern, even when 
the odor of the water is strongly objectionable, because of their 
entire faith in the purity of the original source. 
Thus water from a dirty cistern in West Troy showed.the 
following analysis. In appearance the water was good. 248 
WATER-SUPPLY 
Free ammonia i. 050 per million 
Albuminoid ammonia 175 
Chlorine 2.000 
Nitrogen as nitrites strong trace 
Nitrogen as nitrates 0.0 per million 
Required oxygen '2.25 " 
Total residue 20.00 " 
The roof upon which rain is caught is a twofold cause of 
impurity in the collected water; first, because of the material 
of which the roof is constructed, and, second, because of the 
foreign substances that may settle thereupon. 
In cities the amount of street-dust blown upon the roof 
and afterwards washed into the cistern is much greater than 
is commonly supposed. Soot, excrement of birds (often a 
large item), fallen leaves, horse manure, and various mossy 
growths are among the sundry additions to be found in a roof- 
collected water. 
A question of the first importance in considering a rain- 
water supply is the material out of which the walls of the 
storage-cistern are to be made. Slate and stoneware naturally 
suggest themselves as the most suitable materials, but they are 
not always available, especially if the cistern be a large one 
Cement linings, particularly for underground structures, are by 
far the most common, and the objection that the lime in them 
may somewhat increase the hardness of the water is not of 
much weight, in view of their convenience and low cost. 
Tanks of wood serve their .purpose well, provided they be 
kept full; but if there be great fluctuation in the water-line, 
organic development is liable to occur, and the tank itself falls 
out of repair. The city of New Orleans formerly possessed 
many tanks of cypress-wood. 
Cisterns of metal are open to a number of objections. Iron 
rusts and colors the water; lead is dissolved by rain-water very 
energetically, and is consequently highly objectionable; zinc 
is attacked, and so also is galvanized iron. Tin would be a RAIN, ICE AND SNOW 
249 
suitable metal, but pure tin would be too expensive, and " tin- 
plate " is often defective and is open to rust. 
When the controlling circumstances demand a metallic- 
lined cistern, the metal chosen should be thoroughly coated 
with a good asphaltum paint, or else heavily tinned copper 
should be used. 
The commonly employed delivery-pipe which dips into, 
and remains in permanent contact with, the water of the 
cistern, should also be coated w'ithin and without like the 
cistern walls. 
It is exceedingly important that every cistern should be 
inspected and cleaned frequently, and upon no point does the 
public require more instruction than this. 
One form of underground cistern which has been very 
widely favored in the past is that belonging to the " filtering " 
type. It is constructed by simply dividing the cistern into two 
chambers by a vertical brick wall. Water enters one of these 
divisions and is drawn from the other after percolation through 
the dividing wall. Such an arrangement cannot be too strongly 
condemned. The wall is a mere strainer at the best; it can- 
not be properly cleaned, and it gives a false sense of security. 
The suitable location of an underground cistern is a matter 
that one might think could be safely left to the good sense of 
the average householder; but such is far from being the fact. 
The writer examined one case in which, on account of a defec- 
tive lining and a leaky sewer, a portion of the house-drainage 
was returned to the house along with the cistern-water and 
used for household purposes. In another instance an unlined 
cesspool was observed located in a bank ten feet above and 
fifteen feet to the west of the well furnishing the family's supply 
of water. 
Wherever the cistern is located it should be protected from 
mosquitoes, by screening if necessary. The wooden roof- 
tanks so often seen on buildings supply excellent breeding 
places for these pests. 
While its softness recommends it for use in the laundry, 250 
WATER-SUPPLY 
and while the absence of lime-salts renders it desirable for 
cooking, rain-water is, on the whole, not to be considered so 
suitable as a high-grade ground or surface water for general 
domestic supply.* 
Ice, especially in America, is unquestionably to be ranked 
as an article of food. 
Throughout the colder sections of the country " natural 
ice " controls the market almost completely, and some of the 
dealers supplying the same " harvest their crop " from the first 
sheet of water they find conveniently located, without inquiry as 
to the suitability of its condition, assuming that the process of 
freezing eliminates all objectionable features that the water may 
chance to possess. The author has examined ice from ice- 
houses deriving their supplies from canals, barnyard ponds, 
and the like-localities from which no one would ever dream 
of drawing a supply of water. 
The law of Massachusetts, enacted in 1886, to prevent the 
sale of impure ice reads: 
" Upon complaint in writing of not less than twenty-five 
consumers of ice which is cut, sold, and held for sale from any 
pond or stream in this commonwealth, alleging that said ice is 
impure and injurious to health, the State Board of Health may 
appoint a time and place for hearing parties to be affected, and 
give due notice thereof to such parties, and, after such hearing, 
said board may make such orders concerning the sale of said 
ice as in its judgment the public health requires." 
In New York the law is: 
11 No ice shall be cut for domestic, family, hotel, boarding 
house, restaurant, or other similar use, or for sale for any such 
use, from any impure or contaminated source upon any waters 
within this state. No person, firm or corporation shall cut 
or harvest any ice on any of the waters within this state for any 
such use without first obtaining from the local board of health 
having jurisdiction over the waters upon which it is so cut 
or from the state commissioner of health a written permit. 
* For a description of the water supply of Gibraltar see p. 349. RAIN, ICE AND SNOW. 
251 
A permit granted by a local board of health may be revoked 
by such board or by the state commissioner of health. A local 
board of health or the state commissioner of health may in- 
clude in such permit such instructions and regulations regarding 
the manner in which such ice shall be cut and harvested as may 
be deemed necessary to protect the public health, and the 
state commissioner of health may by general or special rules 
prescribe such regulations. A certified copy of any permit 
issued by a local board of health shall immediately after .its 
issue be filed by such local board of health in the office of the 
state commissioner of health, and such permit shall not be 
effective until so filed. No ice cut from any impure or con- 
taminated source or dangerous to health shall be sold or offered 
for sale for any domestic, family, hotel, boarding house, res- 
taurant, or other similar use. The state commissioner of health 
is hereby authorized to make such rules and regulations and the 
said commissioner and local boards of health are authorized 
to make such inspections as may be necessary to enforce the 
provisions of this section. Any person who cuts, harvests, 
sells, or offers to sell, or stores or transports with intent to 
sell, any ice contrary to the provisions of this section shall be 
liable to a penalty not exceeding five hundred dollars. Each 
day that ice is cut contrary to the provisions of this section 
shall constitute a separate offense. 
Reference has already been made to the small quantity of 
purification to be expected from the freezing of water when 
judged by chemical standards: and Dr. Prudden has shown 
how imperfect the result is when viewed as a bacteriological 
question*. 
He found alternate freezing and thawing more fatal to 
bacterial life than a more prolonged period of continuous 
freezing. 
The prevalent belief that water completely purifies itself 
during the act of freezing is so well fixed in the public mind 
that the source whence ice is harvested is in consequence usually 
* Medical Record, March 26, 1887. 252 
WATER-SUPPLY 
considered unimportant. This belief is not without some sup- 
port in fact, but like many another partial truth it is a very 
uncertain foundation upon which to erect the best form of sani- 
tary procedure. 
A cake of ice in forming thickens by additions to its bottom; 
in other words, it grows downwards. As the crystals enlarge 
they tend to shove out and away from themselves all forms of 
foreign matter which may be suspended in the water, thus puri- 
fying the ice layer and correspondingly increasing by concen- 
tration the pollution of the water below. The amount of 
such concentration in a river would, of course, be only nominal. 
Should the ice be formed upon a shallow basin, however, the 
degree of the concentration might become serious, and if the 
water should freeze throughout its entire depth all of the 
impurities present would of necessity be found in the ice and 
the major part of them would appear in the portion last frozen. 
This state of things may be well seen in a cake of artificial 
ice made from impure water. All the suspended foreign mat- 
ter will be observed collected at the centre of the cake, where it 
was driven and finally entangled by the ingrowing crystals which 
began their formation at the periphery. 
Certain experiments were undertaken in this laboratory 
by Messrs Sherry . and Morey upon water containing 64,000 
bacteria per cubic centimetre. 
When a ten-gallon sample of this water was allowed to 
freeze solid, stored for three weeks and the ice then sampled 
from the centre of the cake, where concentration was greatest, 
such ice was found to contain only 72 bacteria per cubic centi- 
metre, showing a purification efficiency of 99.87 per cent. 
In the center of another cake (from water containing 40,000 
per cubic centimetre) about one quart of water yet remained 
unfrozen at the end of a week and this water (containing 4000 
per cubic centimetre) exhibited a purification efficiency of 
only 90 per cent. This latter figure represented the result 
of a reduction in bacteria due to certain of the freezing condi- 
tions and an increase in the count due to concentration. At 
the end of the second week, the entire cake having meanwhile RAIN, ICE AND SNOW 
253 
frozen solid, the ice of the centre showed 400 bacteria per 
cubic centimetre, at the end of the third week 140, and at the 
end of the fifth week 125. 
Sparks found that in stored ice " The colon bacilli were 
reduced in number about 65 per cent in two days, 90 per 
cent in four, 95 per cent in one week, 99.3 per cent in two weeks, 
and 99.99 per cent in nine weeks, and sterility was found in 
twelve weeks. Typhoid bacilli were reduced in numbers 72 
per cent in two days, 91 per cent in four, 94 per cent in one 
week, 96 per cent in two weeks, 99.99 per cent in seven weeks, 
and sterility was found in nine weeks." 
According to Sedgwick and Winslow " Less than 1 per cent 
of the typhoid germs present in water can survive fourteen 
days of freezing, and that during the first half hour of freez- 
ing a heavy reduction takes place amounting to perhaps 50 per 
cent; after this the reduction proceeds pretty regularly as a 
function of time." * 
Much has been done in the matter of investigating the 
beneficial effects of the freezing action and valuable data will 
be found in the Report of the Massachusetts State Board of 
Health, notably in the volume for 1902, page 271. There is 
undoubted advantage to be secured by ice storage, and without 
question the summer supply taken from the ice-houses is safer 
than that recently cut during the winter 
As Sedgwick has aptly said, however, " it is small comfort 
to the individual suffering from typhoid fever contracted from 
polluted ice to be told that 99 per cent of his friends use ice with 
impunity." 
Ice-borne epidemics of typhoid are fortunately rare, but not 
unknown, as is instanced by the outbreak of the fever at the 
St. Lawrence State Hospital, near Ogdensburg, N. Y.f 
It is not an unknown practice, although a very unworthy 
one, for icemen to quickly build up their " crop " by flooding 
the thin ice-field with ri ver-water. This is commonly done 
to wet down a newly fallen layer of snow and thus avoid the 
* Am. Acad. Arts and Sci., xii. No. V. 
f Am. J. Med. Soc., Oct., 1903. 254 
WATER-SUPPLY 
expense of scraping the ice-field. As a result the cake grows 
in both directions; but there is no opportunity for natural 
purification to take place in the upper or flood portion, which 
part must of necessity freeze as a whole, thus retaining in the 
ice everything which may have been present in the water. 
With reference to the degree of purification taking place 
upon the under side of the ice-cake, it is, as already stated, a 
matter of decided importance, but it is nevertheless short of 
what is claimed for it by most of our people. 
As the crystals of ice form they are sure to entangle more 
or less suspended material, particularly that which is specifically 
light and which therefore tends to hug the under side of the 
growing cake. 
Hence the fact that melted ice often shows more vegetable 
debris than the water from which it is formed. 
Bacteria, whether of harmful form or not, are minute 
vegetable entities which, though invisible to the eye, are never- 
theless suspended in the water and are subject to the same laws 
of exclusion or entanglement by the ice-crystals as are the 
particles of larger and visible size. 
So far as damaging the appearance of the ice is concerned, 
bacteria are of no account whatever, but it is wise to repeat 
here that no reliance must be placed upon the conditions of 
ice formation to produce complete removal of the objection- 
able organisms. 
Judging from the results of an investigation by the writer 
of ice from the upper Hudson district, such removal under 
large scale commercial conditions, and without allowing for the 
benefit of storage, would appear to lie between 90 and 95 per 
cent of all bacteria present in the water.* 
This percentage of purification would be increased through 
the influence of storage, but some of the germs that we are 
anxious to eliminate might remain among the number retained 
by the ice. 
It seems evident, therefore, that if pathogenic bacteria 
become frozen into the ice-cake there is possibility that some of 
* See page 256. RAIN, ICE AND SNOW 
255 
them may reach the consumer in a living condition, and capable 
of doing mischief, even though a period of months should 
elapse between the " harvesting of the crop " and the delivery 
for consumption. 
Of course the ideal rule to follow is never to cut ice from 
a source whence it would be undesirable to drink the water, 
but like many other ideals this one presents practical difficul- 
ties in the way of its attainment. 
In view of the very large imvestment represented by the ice 
industry it might be unwise and unfair to suddenly impose too 
severe conditions upon those engaged in the ice business, but it 
does seem reasonable to insist that the ice furnished the people 
for use in beverages should be required to show negative results 
when tested for the presence of the Bacillus coli communis. 
As concerning the relative merits of transparent and snow 
ice, Prudden gives the following determination of bacteria per 
cubic centimetre contained in the two varieties of ice cut from 
the same cake: 
Bacteria per Cubic Centimetre 
in the melted ice. 
| Transparent ice : 46 
[ Snow-ice 10,020 
f Transparent ice 3,192 
[ Snow-ice 15,624 
J Transparent ice 2,322 
[Snow-ice 55,062 
( Transparent ice 218 
[ Snow-ice 9690 
| Transparent ice 918 
[ Bubbly-streak ice 26,049 
The white ice is richer in bacteria, because it contains large 
quantities of air, and therefore is capable of more readily sup- 
porting the aerobic varieties. It is also exposed to• the im- 
purities likely to be introduced by the " flooding " practice 
already alluded to. The bacteria have not been excluded during 
freezing. 256 
WATER-SUPPLY 
ANALYTICAL RESULTS OF A CHEMICAL AND BACTERIOLOGICAL EXAMINATION OF SAMPLES OF ICE AND THE WATER 
FROM WHICH IT WAS FROZEN. (PART OF REPORT SUBMITTED TO MANUFACTURERS' ASSOCIATION OF NEW YORK.) 
Derivation of Samples. Remarks. 
Results in parts per million. 
Date, 1901. 
Sample. 
Color, 
Platinum Scale. 
Sediment, 
Quickly Falling. 
Turbidity, per 
Clay Standard. 
Free Ammonia. 
Albuminoid 
Ammonia. 
Chlorine in 
Chlorides. 
Nitrogen in 
Nitrates. 
Nitrogen in 
Nitrites. 
Qxygen required 
to Oxidize Organic 
Matter at 2120 F. 
Alkalinity (as 
parts of CaCOa) 
Total Solids. 
Bacteria per 
Cubic Centimetre. 
Bacillus coli 
communis. 
From Ice-house 18, Map I, Mohawk Basin. 
Taken from ice-field during harvesting. 
Water about nine feet deep 
This ice was fifteen inches thick and was clear. 
Feb. 19 
Water 
Ice 
• IS 
None 
Slight 
Slight 
4 
None 
.277 
. 041 
. 103 
.093 
5-5 
i-5 
■ 332 
Trace 
.005 
None 
7-95 
.8 
52.5 
None 
152 
13 
4S0 
40 
Present 
None 
From Ice-house 21, Map 1, Mohawk Basin. 
Taken from ice-field during harvesting. 
Water about twenty feet deep. 
The melted ice was, in appearance, inferior 
to the water. It contained considerable 
floating vegetable debris. 
Feb. 21 
Water 
Ice 
■ 15 
Trace 
None 
See Re- 
marks 
4 
4 
. 282 
• 095 
.098 
. 023 
6.5 
I. 
• 5 
Trace 
.006 
.001 
8.07 
1.8 
50. 
None 
180 
5 
1.440 
150 
Present 
None 
From Ice-house 11, Map 1, Upper Hudson, 
above Mohawk Junction (middle of river). 
Water eight feet deep. 
The ice was twelve inches thick and much 
superior in appearance to the water when 
melted. 
March 1 
Water 
Ice 
3 
None 
None 
Marked 
IS 
18 
.047 
.158 
.067 
. 135 
I. 
• 15 
Trace 
. 0005 
Trace 
31-3 
5-3 
25 • 
None 
118 
103 
2,100 
42 
Present 
None 
From Ice house 16, Map 1, Hudson River just 
above State dam. Water was 13 feet deep. 
The ice was ten inches thick, clear, and very 
fine in appearance. 
March 5 
Water 
Ice 
• 25 
None 
None 
None 
18 
2 
. 168 
.04 
.135 
. 069 
2 . 
I . 
. 219 
.019 
. 0015 
None 
17. 
I . 2 
35- 
None 
139 
20 
4,860 
70 
Present 
None 
From Ice-house 23, Map 1, Back channel of 
Hudson River, below Troy. Water five 
feet deep. 
Feb. 27 
Water 
Ice 
None 
None 
Consid- 
erable 
18 
20 
.327 
. 282 
■ 448 
• 258 
10. 
3-5 
.875 
.112 
.*02 
. 002 
10. 
5-9 
62.5 
None 
247 
128 
91,000 
120 
Present 
None 
From midway between Ice-houses 4 and 5, 
Map 2, below Lower Bridge, Albany. Lo- 
cation exposed to Albany sewage. 
Feb. 22 
Water 
Ice 
.3 
None 
Slight 
Slight 
6 
4 
■ 125 
. 081 
• 147 
.034 
5- 
1 25 
.25 
.125 
. 002 
. OOI 
16. 
2 . 
35- 
None 
149 
24 
5.360 
570 
Present 
Present 
From Erie Canal near Ice-house 15, Map 1, 
used to fill ice-pond. 
Appearance of melted ice inferior to the water 
Feb. 25 
Water 
Ice 
. IS 
None 
None 
Consid- 
erable 
6 
25 
. 298 
. 138 
. 102 
. O98 
6. 
1.75 
.4 
. 007 
None 
6.8 
2.6 
52 5 
None 
188 
75 
2,000 
60 
Present 
None RAIN, ICE AND SNOW 
257 
The slower the formation of ice, and the deeper the water 
on which it forms, the better will be its quality, other things 
being equal. The rate of formation decreases as the ice thickens, 
hence it follows that the lower portion of a thick layer is purer 
than the upper. 
At 140 F. ice forms at following rate: * 
1 inch in 2.06 hours 
6 " 1.95 days 
10 " 5.19 " 
1 foot in 7.38 " 
2 " 28.60 " 
3 " 63.69 " 
It may not be out of place to interject here the following 
army rules relating to the strength of ice. 
Two-inch ice will sustain a man or properly spaced infantry; 
four-inch ice will carry a man on horseback or cavalry or light 
guns; six-inch ice, heavy field guns, such as 8o-pounders; 
eight-inch ice, a battery of artillery with carriages and horses, 
but not over 1000 pounds per square foot on sledges; and ten- 
inch ice sustains an army or an innumerable multitude. On 
fifteen-inch ice railroad-tracks are often laid for months, and 
ice two feet thick withstood the impact of a loaded passenger- 
car after a sixty-foot fall (or perhaps 1500 foot-tons), but 
broke under that of the locomotive and tender (or perhaps 
3000 foot-tons). 
Quarrying ice from the Swiss glaciers has been practiced 
upon a commercial scale, the ice having been broken out by 
blasting. 
Ice in very notable quantity occurs in various parts of the 
world in the so-called " ice-caves " or "glaciers." These 
caves are sometimes of considerable size. Thus, according to 
Balch, the cave at Dobsina, in the Carpathians, is about 10 
metres high, 120 metres long, and some 40 metres broad. 
" The stalactites form grand ice-pillars. They are from 
* See Barnes, " Ice Formation," page 103. 258 
WATER-SUPPLY 
8 to 11 metres in height, and from 2 to 3 metres in diameter. 
In some of the caves, as at Chaux-les-Passavant, the ice stal- 
agmites take nearly the form of cones. There are some seven 
or eight of them, the tallest of which is at least 6 metres high 
with a diameter at the bottom of 5 to 6 metres. Sometimes 
these cones are hollow, as is the case in a grand one at the 
Schafloch, some 6 metres or more in height." 
" This great cave was entirely cleared of its ice in 1727 by 
the Duc de Levi for the use of the Army of the Saone. In 
1743 the ice was formed again. At Szilize, every year, the 
ice has almost completely disappeared by November, and the 
cave is free; but in April or May the floor is again covered 
with ice, and columns and icicles have formed on the roof and 
sides. At La Genolliere the cave is used by the people of 
the neighboring chalets through the spring and early summer 
to help in the operation of butter-making." 
Cave ice is built up from the freezing of percolating water. 
"It is formed entirely by the cold of winter; the heat of 
summer tends to melt it. Owing to the sheltered position of 
cave ice the summer heat reaches it with more difficulty than 
it reaches the snow and ice in the open, and it therefore remains 
long after the ice in the surrounding country has disappeared." 
Anchor ice, while not a source of supply, may deserve a 
word here for the reason that it not uncommonly becomes 
troublesome by blocking up the entrances of water-works 
intakes. 
Heat radiation from submerged surfaces, especially those of 
dark color, will in higher latitudes often allow of ice formation 
beneath the water-level, as upon the intake strainers, and this 
ice once started may be greatly added to by accumulated masses 
of " needle " or " frazil " ice mechanically carried down by 
descending water currents, and then frozen to the anchor ice 
proper. Cold, clear nights are favorable to these forma- 
tions. During the day masses of this submerged ice will 
occasionally loosen and rise to the water-surface, but that por- 
tion which is attached to the irregularities of the intake-screen RAIN, ICE AND SNOW 
259 
is hard to detach. Steam and warm air have been used for its 
removal but such a remedy is expensive and may not be avail- 
able in remote localities. An electric current through the 
strainer bars has been used with good results. A simple device 
for avoiding anchor ice formation, one which is always worth 
trying but which is not uniformly successful, is to build a raft 
or platform over the site to be protected; the object being to 
prevent radiation of heat from the surfaces beneath. Of 
course a covering of sheet ice would serve the same purpose. 
It is claimed that anchor ice does not form at depths below 
forty feet. It has never troubled the intakes of Toronto or 
Milwaukee, which draw from depths of eighty and forty-six 
feet respectively. 
Artificial ice is making very rapid strides toward popularity, 
and if its manufacturers would confine themselves to the use of 
thoroughly safe water as a basis for their product, there is no 
question but that the confidence of the people would be well 
placed and permanently retained. 
Unfortunately, there is a quantity of artificial ice offered for 
potable use that is made from very ordinary water; and, 
inasmuch as the usual method of formation causes the water 
used to freeze as a whole, all the impurities of the water are 
retained and concentrated in the centre of the cake of ice, 
that being the last portion to solidify. Unless the water 
employed be distilled, such artificial ice must, of necessity, 
be more impure than natural ice frozen from the same water. 
The objection advanced above does not, however, apply to 
a method of preparing ice whereby the processes of nature are 
closely imitated, the ice being permitted to form on a vertical 
freezing plate immersed in flowing w'ater. 
Snow can be considered only as an indirect source of water 
supply, but as such it assumes a position of some importance. 
The water from melted snow (which, according to the U. S. 
Weather Bureau, averages in depth one-tenth that of the 
snow before melting) is commonly more impure than rain- 260 
WATER-SUPPLY 
water from the same locality, for the reason that its flakes 
act better than the spherical rain-drops for entangling im- 
purities suspended in the atmosphere, and their low tempera- 
ture is conducive to the absorption of ammonia. Analysis 
of city and country snows from the same general locality show 
marked differences which are illustrated in the results obtained 
from samples gathered in the city of Troy, N. Y., and in the 
near-by open country: 
City Snow. 
(Troy, N. Y.) 
Country Snow. 
(Menands Station.) 
Free ammonia 
Albuminoid ammonia 
Nitrogen as nitrates 
Nitrogen as nitrites 
Chlorine 
Required oxygen 
Parts per million. 
.460 
.225 
. 200 
Trace 
1.87 
1.90 
Parts per million. 
•15 
.06 
Trace 
Slight trace 
.60 
1.00 
Of course, as is the case with rain, the first portion of the 
fall must always contain the greatest amount of impurities. 
The chlorine in city snow (Troy, N. Y.) was thus found to 
vary during the same storm of two days' duration. (Not the 
same storm as above.) 
First day 3.05 parts per million 
Second day 2.55 
London snow was found by Coppock * to contain: 
Total solids 237.3 per million 
Mineral matter 89.3 " " 
Carbonaceous matter 156.5 " 11 
Free ammonia 66.3 tc " 
Albuminoid ammonia 93 . o " c c 
The first half of the above-referred-to snowfall contained 
75 per cent of the impurities. The carbonaceous matter was 
ordinary soot. 
* Chem. News, Ixxi. 92. RAIN, ICE AND SNOW 
261 
After snow is once upon the ground it changes in composi- 
tion quite rapidly, particularly in its contained ammonia. This 
change is, moreover, greatly influenced by the character of the 
surface upon which it rests. 
Thus the tendency of snow to absorb impurities from the 
soil is shown by the following comparative analysis of samples 
taken from a roof and from a meadow (results in parts per 
million): 
Free 
Am- 
monia. 
Albu- 
minoid 
Am- 
monia. 
Chlor- 
ine. 
Nitro 
gen as 
Ni- 
trites. 
Nitro- 
gen as 
Ni- 
trates. 
Re- 
quired 
Oxy- 
gen. 
Total 
Resi- 
due. 
Loss 
on Ig- 
nition. 
Fresh snow from roof.. 
Same snow after lying 
•5° 
•15 
. 80 
trace 
trace 
•5° 
22. 
7- 
on roof two days.. . . 
Fresh snow from mead- 
1.24 
. 21 
•85 
trace 
trace 
2.85 
64. 
21 . 
ow 
.27 
.11 
•75 
trace 
trace 
•40 
41. 
12. 
Same snow after lying 
in meadow two days. 
•79 
. 26 
.70 
trace 
trace 
2.70 
65- 
29. 
Additional force is thus given to the saying, " Snow is the 
poor man's fertilizer," and " The fogs and snow remain to 
fatten the land." 
The great influence that melting snow has upon spring- 
water is shown by the following analyses of a flow from a 
spring in Rensselaer County, N. Y. The water serves as an 
illustration, although it is not a " normal " water, as is seen 
from the high chlorine. 
Results in parts per 
million. 
Oct. Nov. 
15- io. 
Dec. 
15. 
Jan. 
12. 
Feb. 
5- 
Mar. 
2. 
April 
6. 
May 
8. 
June 
5- 
Free ammonia 
Albuminoid ammonia. 
Nitrogen in nitrites. . . 
Nitrogen in nitrates.. . 
Chlorine 
trace trace 
.036 .03 
.000 .000 
.116 .075 
20.S 17.3 
.000 .1 
570. 570. 
79. 
53-6 49. 
158. 750. 
.015 
.055 
. 000 
. 15 
19. 
•4 
558. 
62. 
46.4 
1620. 
. 010 
.035 
. 000 
• 15 
19. 
1.15 
534- 
44. 
44-6 
2519. 
.025 
. 090 
trace 
trace 
21. 
i-5 
579- 
114. 
4i • 
166. 
. 027 
.078 
. 000 
.30 
20. 
I . 0 
543. 
73- 
43-7 
8520. 
.025 
. 060 
. 000 
. 15 
21.5 
• 3 
554- 
63- 
42. 
476. 
.01 
■04 
. . 000 
trace 
18. 
. 1 
553. 
48. 
50. 
.01 
■07 
.000 
• 05 
17. 
• 5 
552. 
52. 
51.8 
"Required oxygen''.. . 
Total solids 
Loss of ignition 
Temperature F 
Bacteria per c.c CHAPTER VI 
RIVER AND STREAM WATER 
A very large number of cities derive their water-supply 
from rivers-in Europe, after careful filtration, but in America 
often without such purification. 
One of the important things for the consumer of a river- 
water to bear in mind is that sudden and great changes in its 
character are to be expected. 
For instance, the water of the Hudson River, sampled at a 
point above direct sewage-inflow (although below several large 
towns), showed the following variations: 
HUDSON RIVER 
(Parts per million) 
2 "0 B m p 
■ j* o & 5: o 
M cn 00hCr In' 
: o ; ; p : 
n p £ 5° 
• 030 
• 045 
• 025 
•055 
• 085 
.042 
• 058 
• 030 
• 045 
. 280 
■055 
Free 
Ammonia. 
.087 
■150 
.080 
. 100 
• 150 
•235 
.660 
• 205 
.120 
• 320 
•i55 
Albuminoid 
Ammonia. 
000 
trace 
000 
000 
trace 
trace 
trace 
trace 
000 
■ 0015 
trace 
Nitrogen as 
Nitrites. 
trace 
-*5 
. 10 
• i5 
. 10 
■30 
trace 
. 10 
trace 
•30 
Nitrogen as 
Nitrates. 
3 • 5 
4-5 
3-5 
2-4 
2-5 
3-5 
5-o 
3-5 
Chlorine. 
3-76 
13-00 
7-65 
8.85 
10.00 
5-9° 
I5-5O 
7-3° 
8.70 
2-65 
14-85 
"Required 
Oxygen." 
73- 
107. 
43- 
88. 
93- 
388. 
583- 
67. 
78. 
101. 
Total 
Residue. 
35- 
42- 
39- 
45- 
5i. 
88. 
74- 
3i- 
,s?: 
57- 
Loss on 
Ignition. 
88.4 
68.8 
000 
000 
000 
11. 
495- 
000 
000 
000 
Suspended 
Matter 
(Silt.) 
36. 
34-6 
33- 
46.4 
41. 
68. 
71. 
44- 
Temperature 
F°. 
A ri ver-water which is clear to-day may be muddy and less 
desirable for use to-morrow. 
Fischer gives the following seasonal variations for the water 
of the Danube (in parts per million): 
262 RIVER AND STREAM WATER 
263 
Suspended 
Material 
Dissolved 
Material 
Spring 
121.9 
177.I 
Summer 
165.4 
146.O 
Autumn 
76.5 
178.6 
Winter 
14.8 
199.O 
It is to be noted that the bulk of variation lies in the item 
of suspended material, and that what is in solution is much 
more constant in amount.* 
The Gila river at San Carlos, Arizona, is probably the most 
muddy in the United States. During the period July 21st 
to October 24th, 1904, its turbidity varied from 220 to 73,536 
parts per million.f 
Some idea of the immense quantities of materials carried 
to the sea by river-flow may be gathered from the fact that 
the Colorado River discharges per year 338,000,000 tons of 
suspended silt, in addition to 4,550,000 tons of common salt, 
3,740,000 tons of Glauber's salts, 4,000,000 tons of lime, 
2,400,000 tons of gypsum and 4,800,000 tons of Epsom salts 
all held in solution.^ 
C. H. Stone, of the U. S. Reclamation Service, reports 
the following analysis of Mississippi River water § taken May 
23, 1905, opposite Nine-Mile point just above Carrollton, in 
midstream, and about six feet below the surface. 
(Expressed in Parts per Million) 
Total solids (unfiltered) 1069. 
Total solids (filtered) 
Loss on ignition (unfiltered) 74. 
Loss on ignition (filtered) 27.5 
Si 3-5 
Al , 0.09 
Mn 0.12 
* For sundry American data by E. G. Smith see J. Am. Water-works Asso. 
1896, p. 88. 
f Water Supply and Irrigation Paper No. 151. 
I U. S. Geol. Sur. Water-Supply Paper 274. 
§ Science, Oct. 13, 1905. 264 
WATER-SUPPLY 
Ca 29.5 
Mg 6.8 
Fe 0.08 
K 2.3 
Na 10.o 
SO4 28.7 
PO4 0.4 
CO3 0.0 
HCO3 110.4 
Cl 16.1 
Nitrogen as free ammonia 0.16 
Nitrogen as albuminoid ammonia 0.14 
Nitrogen as nitrites 0.0 
Nitrogen as nitrates 0.23 
Oxygen consumed (unfiltered) 14,2 
Oxygen consumed (filtered) 3.3 
Hardness 109.2 
Turbidity Heavy 
Sediment Large 
Odor (cold) Practically none 
RESULTS OF ANALYSIS CALCULATED AS OXIDES 
SiO2 7.4 
AI2O3 0.17 
Fe2Os . o. 11 
Mns04 0.16 
CaO 41.2 
MgO 11.3 
K20 2.8 
Na20 13.5 
SO3 23.9 
CO2 79-6 
The silt from the above water sample was saved and reported 
upon later by the same chemist.* 
Referring to his analysis he says' 
* Science, April 20, 1906. RIVER AND STREAM WATER 
265 
"For the sake of argument let us assume that the above 
analysis represents the average composition of the silt carried 
by the Mississippi during the entire year. This is doubtless 
not quite true, but will serve as a basis for some calculations. 
One estimate of the total amount of silt carried by the Mississippi 
during a year places the figures at 443,750,000 tons. Assuming 
this to be true, the following table gives in tons the amounts 
of various substances removed in this silt during the year: 
Fe.Os+AbOs 79,431,250 
MnO 798,750 
CaO 8,120,625 
MgO 7,277,500 
Na O 576,875 
K20 5,591,250 
P2O5 i,io9,375 
SO3 1,142,500 
Total N 665,625 
Water and organic matter 31,062,500 
" For most of these substances it is impossible to assign any 
definite commercial value, but for four of them it is possible to com- 
pute the actual cost of restoring them to the soil in the form of fer- 
tilizer. In the following table such calculation has been made. 
It has been assumed that the potash would be bought in the 
form of K2SO4, the phosphoric acid as superphosphate and the 
N as NaNOj. The figures are of course not absolute, but they 
convey a good idea of the loss which the land has sustained. 
" Value of plant food removed in silt by the Mississippi 
River during one year: 
CaO $40,603,125 
K20 559,125,000 
P2O5 110,937,500 
N 222,984,375 
" These figures are stupendous and worthy of careful con 
sideration, and when we consider that this same process of 
denudation of the land is being carried on by all the streams of 
the country, to a greater or less extent, we gather some faint 
idea of the loss to agricultural interest from this cause." 266 
WATER-SUPPLY 
The surface of the United States is being removed at the 
rate of thirteen ten-thousandths of an inch a year, or one inch 
in 760 years, according to the United States Geological Sur- 
vey Though this amount seems trivial when spread over the 
surface of the country, it becomes stupendous when considered 
as a total, for over 270,000,000 tons of dissolved matter and 
513,000,000 tons of suspended matter are transported to tide- 
water every year by the streams of the United States.* 
River-waters contaminated with unusual materials are at 
times met with. Thus the well-known Rio Vinagre of South 
America contains 1100 parts of free sulphuric acid and 1200 
parts of free hydrochloric acid per million of water. The 
quantity of free sulphuric acid carried by it to the sea is over 
fifty tons daily. 
In the Youghiogheny River near McKeesport, Pa., the 
average free acid as H2SO4 from Sept. 1, 1906, to Sept. 1, 1907, 
as given in Water Supply Paper 236, U. S. Geol. Survey, was 
22 parts per million. From Oct. 1st to Dec. 1st, 1908 the aver- 
age was 140 parts. 
" The variations in the condition of the water are wide 
and sudden. On one occasion the acidity increased from 50 
to 210 parts per million within three hours." 
Some streams of Norway and Sweden furnish water so 
impregnated with infusion of woody material as to be destruc- 
tive of fish. Frankforter found 11 tannates " and " gallates " 
in the water of the upper Mississippi, due to the enormous 
number of logs floated down the stream from the great forests 
of the North. 
Judged also the the bacteriological side, flowing water 
will always show large variation in composition at different 
times, principally due to introduction of impurities carried down 
by storms from surface sources. Variation will also be noted 
at different points of the length, breadth, and depth of the 
same stream, as common judgment would expect, arising from 
* See Science, Dec. 13, 1912. RIVER AND STREAM WATER 
267 
irregularities in mixing of tributary waters, and from changes 
in the rate of sedimentation. 
Tidal action has also much influence upon the variation in 
character of certain river-waters. The author has in mind 
several cities, situated upon large streams, which pump fairly 
good water during ebb-tide, but whose sewage is carried up- 
stream by the reversed current of flood-tide, to and beyond 
the intakes, with exceedingly bad results. 
All such points are to be considered when selecting the 
position for a city's intake. 
The pollution of rivers by the progressive introduction of 
sewage and industrial wastes causes a change that is a more 
serious matter than those seasonal ones noted above; a change 
that, although somewhat slow in growth, has the disadvantage 
of being continually on the increase through the enlarging of 
the populations which naturally sewer into the rivers. 
Many of our rivers are very large, but the volume of city 
sewage dumped into them is large as well. It is sometimes very 
great indeed, as for instance the amount that Chicago daily 
turns into the Illinois River or that Philadelphia pours into the 
Schuylkill and Delaware. 
The Hudson River at the city of New York is not to be 
rated as a fresh-water stream, but we note that the raw sewage 
that the Metropolis daily disposes of therein reaches the great 
volume of five or six hundred million gallons. 
Remembering the old rule of law that a riparian owner is 
entitled to receive from his up-stream neighbor a stream-flow 
undiminished in quantity or quality, the question presents 
itself, should the amount of river pollution now existing be 
permitted to continue. 
This question is a very vexed one and it has been answered 
in sundry ways. Extreme partisans on the one side have raised 
outcry against making sewers of our rivers and they insist 
that the waters should be returned to, and maintained in, 
the state of purity in which the discoverers found them. Those 
on the other side, equally extreme, insist that rivers and streams, 268 
WATER-SUPPLY 
being the natural drains of the country, should be used as such 
and should be assigned the duty of carrying to the sea the 
waste materials of which the valley communities need to dispose. 
The author believes that every case of stream pollution 
should be judged upon its own merits, and he is distinctly 
opposed to rules or laws of a blanket character, feeling that 
blanket legislation adjusted to fit one case might do grave 
injustice if applied to another. As a broad proposition large 
streams, especially navigable ones, should be considered as 
draining a country rather than watering it, but nevertheless 
the up-stream community should not so grossly misuse such 
a drain as to make it border upon a nuisance and render it 
difficult or impossible for the down-stream city to use the water 
for general supply even after applying some reasonable and 
modern form of water purification. While an up-stream town 
should not grossly misuse the stream as a drain, they should, 
nevertheless, not be so unfairly treated as to be required to 
convert their sewage into drinking-water before dumping it 
into the river, for the reason that the down-stream community 
should in justice be asked to bear its appropriate portion of 
the general expense and employ suitable methods for the puri- 
fication of its water supply. 
Small streams which are in no way navigable should be 
placed in a different class and more stringent regulations 
should be enforced to safeguard their purity. Let it be remem- 
bered that the " time of flow " between an intake and an 
up-stream point of pollution is the element of importance, 
rather than the distance in miles, for the reason that a rapidly 
moving brook may deliver its load of objectionable material 
from a distant source more quickly, and consequently in a 
fresher condition, than can be done by a slowly flowing stream 
even though the latter draws its pollution from a point nearer 
at hand. The importance of the esthetic side of the question 
should be remembered as well, and the general appearance of a 
stream should be considered when undertaking to determine 
whether or not its water could be damaged by the introduction 
of some proposed form of waste discharge. RIVER AND STREAM WATER 
269 
In short, aside from any consideration of disease develop- 
ment, the introduction of waste material should breed no 
" nuisance." 
Stearns considers that if the volume of the entering sewage 
be greater than of that of the stream into which it empties, 
a nuisance will result; if its volume be less than t|o of that 
of the stream satisfactory results may be counted upon; while 
if its relative volume be between those figures no prediction 
can be made. 
A convenient form of statement, widely used, is that it 
will require a volume of stream-flow equal to a certain number 
of cubic feet per second (i.e., " second-feet ") to safely take 
care of the sewage from 1000 people. In applying this ex- 
pression attention must be had to the character of the stream 
itself, as an opportunity for aeration will increase the efficiency 
of the water to properly care for its dose of sewage, while chances 
for 11 pooling " and " stranding " of the suspended solids will 
have an opposite effect. 
After all is said it must be admitted that the values and 
forms of definition commonly given do not fully cover the 
situation, nor are we released from any obligation to decide each 
case upon its own merits; for a nuisance is not to be attributed 
to sewage discharge alone, nor would the same cause be voted 
a nuisance by all people. Of two streams with which the writer 
is familiar the first has the objectionable odor due to waste 
liquor from a sulphite-pulp mill while the other is stained 
bright pink by the discharge from dye-works. 
Two men, one blind and the other suffering from loss of the 
sense of smell would give very different interpretations of 
nuisance as applied to these waters. 
In New York law, a nuisance is defined as an act or omission 
which " annoys, injures, or endangers the comfort, repose, 
health or safety of any considerable number of persons; or 
which offends public decency." 
If putrescible organic matter be added to a stream in such 
quantity that the dissolved oxygen is not sufficient to oxidize 
it, then decomposition in absence of oxygen results with its 270 
WATER-SUPPLY 
attendant offensive odors. Proper aeration is the cure for such 
a condition but is must be noted that while aeration disposes 
of the annoyance due to the organic matter it does not destroy 
germ life. 
The killing of fish is one of the objectionable results which 
follow the undue fouling of a stream by sewage or industrial 
waste; their gills become clogged with fibrous material and they 
are deprived of the amount of dissolved oxygen which their 
well-being demands. Necessarily there is considerable dif- 
ference in the quantity of oxygen which different kinds of 
fish require; thus coarse varieties are able to do wTith a smaller 
percentage of saturation than is needed for trout and other 
fish of delicate organism. 
Speaking generally, fish should be furnished with dissolved 
oxygen to an amount equal to from 30 to 50 per cent of sat- 
uration. , 
Weigelt records the following observations of the effects 
upon fish of waters contaminated with products of industrial 
waste: * 
Contained in One Million Parts of Water. 
Kind of 
Fish Used. 
Observations. 
70 slaked lime, Ca(OH)2 
1000 soda, Na2CO3-ioH2O 
0.5 chloride of lime, CaCl2O 
100 hydrochloric acid, HC1 
100 sulphuric acid, H2SO4 
50 ammonia, NH3 
100 sodic arsenate Na2AsO4 • 12H2O. . . . 
50 mercuric chloride, HgCl£ 
1000 calcium chloride, CaCL 
100 green vitriol, FeSOvyHaO 
50 green vitriol, FeSO4-7H2O 
1000 iron chloride, FeCls 
1000 manganese chloride, MnCl2 
5 carbolic acid, CgH6OH 
1000 soap (unfiltered) 
1000 soap (filtered) 
Trout 
i t 
C i 
( c 
I c 
C ( 
c c 
c c 
( c 
c ( 
c c 
I c 
c I 
c c 
Salmon 
C c 
Dead in 26 minutes. 
After 3 minutes, restless. 
Dead in 3 hours. 
On its side in 4 minutes. 
On its side at once. 
Dead in 47 minutes. 
Strong effect. 
Dead in 54 minutes. 
An effect in 2 hours. 
Dead in 5 hours. 
No effect in 16 hours. 
On its side in 3 minutes. 
Speedy restlessness. 
Restless in 15 minutes. 
Dead in ij hours. 
No effect. 
* " Das Wasser," Fischer, p. 52. RIVER AND STREAM WATER 
271 
H. W. Clark * in his paper on 11 Fish Life and Water Pol- 
lution " finds the following chemicals fatal to fish if present in 
the amounts here stated as parts per million: f 
(NH4)2CO3 55 to 7° 
NH4OH 13 
Na.COs 250 
NaOH 70 
NH4CI 180 had no effect 
Iron in solution was found " very fatal." 
In water very slightly acid with nitric acid " the fish not 
only neutralized the acid but made the water alkaline." 
He found that a one-pound fish used up all the dissolved 
oxygen in one gallon of water at 6o° F. in one hour. 
In the same paper the above author gave the following 
data dealing with the effect of sundry sewage dilutions upon 
fish life: 
In straight sewage, kept well aerated (i.e., containing 50 
to 100 per cent dissolved oxygen), all fish died within a few 
minutes. 
When water and sewage, well aerated, were mixed in 
equal portions some fish lived indefinitely, but with a larger 
proportion of sewage to the water they died in a few 
minutes. The above 1 : 1 mixture was fatal to all except 
robust individuals. The sewage used was freed from suspended 
matters. 
Effluents from successful 11 contact-beds " and " trickling- 
filters," if kept well aerated, sustained fish life indefinitely. 
Mixtures of straight (but strained) sewage and tap water, 
not artificially aerated, could not sustain fish life if the sewage 
were over 10 per cent of the mixture. 
Substituting a " trickling-filter " effluent for the sewage 
the water in the mixture did not have to exceed 25 per cent 
* Eighth International Cong. App. Chem., Sec. Villa, 
f For the action of CuSO4 upon fish see page 321. 272 
WATER-SUPPLY 
to sustain fish life, but with " contact-bed " effluents the water 
had to be at least 50 per cent of the mixture. 
Rideal * calls attention to the manifest truth that while 
the death of fish points towards an objectionable water, the 
fact that they live is by no means a guarantee of purity. He 
adds: " Fish are more affected by muddy water and by chemi- 
cals from factories than by excreta." 
The dissolved oxygen required by the major fish life is almost 
entirely the result of direct solution from the air, but it should 
be remembered that the atmospheric gases are not equally 
soluble in water; the ratio of oxygen to nitrogen in the " boiled- 
out air " is different from the atmospheric relation because of 
the greater solubility of the oxygen. 
With reference to the question of dissolved oxygen a great 
deal of painstaking work was done by the Metropolitan Sewage 
Commission and the data secured form no small part of the 
record in the case of New York vs. New Jersey before the 
U. S. Supreme Court in the matter of the pollution of the waters 
of New York Bay. 
It was of course necessary in that case to recognize the fact 
that oxygen dissolves in sea-water less readily than it does in 
fresh water, the values (at 170 C. and 760 m.m. pressure) 
being, f 
Sea-water 5.6 c.c. per litre 
Freshwater 6.9 
As supplementary to what has already been said regarding 
the self-purification of streams, a word may be added to the 
effect that when the end desired is the prevention of the sewage- 
inflow to a river becoming a nuisance, rather than the preser- 
vation of the potable character of the water, then unques- 
tionably falls and rapids become of material value, for they are 
the means of increasing the supply of dissolved oxygen held 
by the water, and upon the quantity of such oxygen present 
* " Sewage," page 44. 
f Report Metro. Sewer Com., 1910, p. 410. 273 
RIVER AND STREAM WATER 
the capacity of the stream to satisfactorily dispose of a 
nuisance following sewage-inflow largely depends. 
RAINFALL, EVAPORATION, AND FLOW OF STREAMS 
Rainfall is greatest within the tropics and near the sea, 
and it lessens near the poles. In the same general district it 
also tends to increase with elevation above sea-level, but local 
disturbing factors enter to offset this tendency and introduce 
irregularities. The average annual precipitation for the north 
temperate zone is usually estimated at 35 inches. 
The great volume of water represented by a heavy rain- 
fall is not generally appreciated, thus one inch of rain falling 
upon an area of one square mile corresponds to 2,323,200 cubic 
feet of water, or nearly 17,500,000 U. S. gallons. 
Exceptionally heavy rainfalls are not, for our purpose, 
expecially worthy of record;' but it may be interesting to note, 
very briefly that the greatest concentrated rainfall on record 
is reported by the Catholic meteorological station at Manila, 
P. I., as having occurred in Western Luzon. During three 
days the total rainfall recorded at Baguio was 88 inches, 32 
inches of which is reported to have fallen during the first 
twenty-four hours.* 
A further list of very heavy rainfalls may be found in Rafter 
and Baker, " Sewage Disposal," page 134. See also Senate 
Doc. 91, 50th Congress. 
" There is a theory which numbers of people accept as true, 
that great battles have been generally, if not invariably, closely 
followed by storms. 
" This belief is deeply rooted in the popular mind, somewhat 
like the various notions held by many people in relation to the 
effects of the moon's phases upon the weather And it appears 
to be a traditional idea, for the belief that battles cause rain 
was prevalent before the invention of gunpowder." f 
" While the question of rain-making by the use of explosives 
* Engineering News, Nov. 30, 1911. 
f Am. Met. Jour., March, 1802. 274 
WATER-SUPPLY 
was under consideration at Washington the scientists of the 
Department of Agriculture made a thorough investigation of 
the subject, with all the records of the government at their 
command, and the conclusion reached was that there is no 
foundation for the opinion that days of battle were followed by 
rain any more than days when it was all quiet along the 
lines." * 
There is no record showing any relation between rainfall 
and the tremendous battles which have occurred during the 
present great war. 
Regarding the relation of great fires to rainfall, Prof. I. A. 
Lapham, of the U. S. Signal Service, writes of the Chicago fire 
as follows: 
11 During all this time-twenty-four hours of conflagration 
upon the largest scale-no rain was seen to fall, nor did any 
fall until four o'clock the next morning; and this was not a 
very considerable downpour, but only a gentle rain that ex- 
tended over a large district of country, differing in no respect 
from the usual rains. It was not until four days afterward that 
anything like a heavy rain occurred. It is, therefore, quite 
certain that this case cannot be referred to as an example of 
the production of rain by a great fire." 
The following chart and statistics f show the normal rain- 
fall of the United States, and also the same data for separate 
States. 
RAINFALL AND SNOW OF THE UNITED STATES 
Annual and seasonal averages, seasonal variation and cubic miles for each State. 
n 
0, 
D 
3 
California 
> 
§ 
£ 
Ji 
> 
w 
3 
■ 
> 
13 
13 
3 
p 
State. 
52,250 
113,020 
53,85° 
158,360 
103,925 
Area in 
Sauare 
Miles. 
Ins. 
14-9 
i-3 
14-3 
6.2 
4-2 
Spring. 
Ins. 
13.8 
4-3 
12.5 
°-3 
5-5 
Summer. 
Ins. 
IO.O 
2.2 
II .O 
3-5 
2.8 
Autumn. 
Ins. 
14.9 
3-1 
12.8 
11.9 
23 
Winter. 
Ins. 
53-6 
10.9 
50.6 
21.9 
14.8 
Annual. 
Ins. 
i-5 
3-3 
3-9 
40.0 
2.4 
Seasonal 
Variation 
44-2 
19.4 
42.5 
54-9 
24.2 
Cubic 
Miles. 
* Sage, Iowa Weather and Crop Service. 
t Report of the Chief of the U. S. Weather Bureau. 5 a WWWMW 
V NORMAL ANNUAL PRECIPITATION 
(inches) 
U.S. Weather Bureau RIVER AND STREAM WATER 
275 
RAINFALL AND SNOW OF THE UNITED STATES-Continued. 
State. 
Area in 
Square 
Miles. 
Spring. 
Summer. 
Autumn. 
| Winter. 
Annual. 
Seasonal 
Variation 
Cubic 
Miles. 
Connecticut 
4,99° 
Ins. 
II . I 
Ins. 
12.5 
Ins. 
11.7 
Ins. 
ii-S 
Insd 
46.8 
Ins. 
I . I 
3-6 
Delaware 
2,050 
IO. 2 
II .0 
10.0 
9.6 
40.8 
I . I 
i-3 
District of Columbia 
70 
II .0 
12.4 
9-4 
9.0 
41.8 
1.4 
0.04 
Florida 
58,680 
IO. 2 
21.4 
14.2 
9.1 
54-9 
2-4 
5i-o 
Georgia 
59,475 
12.4 
15-6 
10.7 
12.7 
5i-4 
i-5 
48.2 
Idaho 
84,800 
4-4 
2.1 
3-6 
7-o 
17.1 
3-3 
22.7 
Illinois 
56,650 
10.2 
11.2 
9.0 
7-7 
38.1 
i-5 
34-0 
Indiana 
36,350 
11.0 
11.7 
9-7 
10.3 
42.7 
1.2 
24.2 
Indian Territory. . . 
3L4oo 
10.6 
11.0 
8.9 
5-7 
36.2 
1.9 
17-7 
Iowa 
56,025 
8-3 
12.4 
8.1 
4-i 
32.9 
3-o 
28.8 
Kansas 
82,080 
8.9 
11.9 
6-7 
3-5 
310 
3-4 
40.0 
Kentucky 
40,400 
12.4 
12.5 
9-7 
11.8 
46.4 
i-3 
29-3 
Louisiana 
48,720 
13-7 
150 
10.8 
i4-4 
53-9 
1.4 
41.6 
Maine 
33,040 
11.1 
10.5 
12.3 
11.1 
45-o 
1.2 
23.2 
Maryland 
12,210 
11.4 
12.4 
10.7 
9-5 
44-o 
i-3 
8-3 
Massachusetts 
8,315 
11.6 
11.4 
11.9 
11.7 
46.6 
1.0 
5-9 
Michigan 
58,915 
7-9 
9-7 
9.2 
7-o 
33-8 
i-4 
31-3 
Minnesota 
83,365 
6-5 
10.8 
5-8 
3-i 
26.2 
3-5 
34-4 
Mississippi 
46,810 
14.9 
12.6 
IO. I 
15-4 
53-o 
i-5 
38.8 
Missouri 
69,415 
10.0 
12.4 
9.1 
6-5 
38.0 
1.9 
41.2 
Montana 
146,080 
4-2 
49 
2.6 
2-3 
14.0 
2.1 
32.1 
Nebraska 
77,5io 
8.9 
10.9 
4.9 
2.2 
26.9 
5-o 
32.9 
Nevada 
110,700 
2.3 
0.8 
i-3 
3-2 
7.6 
4.0 
14-4 
New Hampshire. . . . 
9,305 
9.8 
12.2 
11.4 
10.7 
44-i 
1.2 
6-3 
New Jersey 
7,815 
II.7 
13-3 
11.2 
11.1 
47-3 
1.2 
5-6 
New Mexico 
122,580 
i-4 
5-8 
3-5 
2.0 
12.7 
4-i 
24-5 
New York 
49A7O 
8-5 
10.4 
9-7 
7-9 
36.5 
i-3 
28.3 
North Carolina 
52,250 
12.9 
16.6 
12.0 
12.2 
53-7 
1.4 
44-2 
North Dakota 
70,795 
4-6 
8.0 
2.8 
i-7 
17.1 
4-7 
19.1 
Ohio 
41,060 
10.0 
11.9 
9° 
9.1 
40.0 
i-3 
25-7 
Oregon 
96,030 
9.8 
2-7 
10.5 
21 .O 
44-0 
7-8 
66.7 
Pennsylvania 
45,215 
10.3 
12.7 
10.0 
9-5 
42.5 
i-3 
30.2 
Rhode Island 
1,250 
11.9 
10.7 
11.7 
12.4 
46.7 
I . 2 
0.8 
South Carolina 
30,570 
9.8 
16.2 
9-7 
9-7 
45-4 
i-7 
21.6 
South Dakota 
77,650 
7-2 
9-7 
3-5 
2-5 
22.9 
3-9 
28.1 
Tennessee 
42,050 
13-5 
12.5 
10.2 
14-5 
50.7 
i-4 
33-4 
Texas 
265,780 
8.1 
8.6 
7.6 
6.0 
30.3 
1.4 
127.0 
Utah 
84,970 
3-4 
1,5 
2.2 
3-5 
10.6 
2-3 
143 
Vermont 
9,565 
9-2 
12.2 
11.4 
9-3 
42.1 
i-3 
6.1 
Virginia 
42,450 
10.9 
12.5 
9-5 
9-7 
42.6 
i-3 
28.5 
Washington 
69,180 
8.6 
3-9 
10.5 
16.8 
39-8 
4-3 
43-4 
West Virginia 
24,780 
10.9 
12.9 
9.0 
10.0 
42.8 
i-4 
16.6 
Wisconsin 
56,040 
7-8 
11.6 
7-8 
5-2 
32-5 
2.2 
28.7 
Wyoming . . . . 
97,890 
4-3 
3-5 
2.2 
1.6 
11.6 
2-7 
17-9 
Total 
2,985,850 
1407.14 
Average 
9.2 
10.3 
8.3 
8.6 
36-3 
3-0 276 
WATER-SUPPLY 
" A very remarkable feature in the rainfall of the United 
States, appearing on most of the monthly maps, and distinctly 
on the annual map, is the way in which certain peaks and 
ranges of mountains are outlined by the mean rainfall. 
" Another series of facts of very great interest can be read 
from the maps in the consideration of the relations of rainfall 
to the leeward and windward sides of the ranges. This is by 
far the best marked on the Pacific coast, where the prevailing 
winds are distinctly from the west and reach the coast laden 
with moisture from the warm ocean. To the westward, for 
instance, of the Sierra Nevadas on the annual map there is a 
rainfall of from 20 to 40 inches. Immediately to the east- 
ward of this series of mountains the annual rainfall is only 
from 2 to 6 inches. Much the same is true of the Cascade 
Range, and even the Coast Range has a very marked influence 
on the rainfall. The annual line, for instance, of 40 inches 
of rainfall passes down the coast from Vancouver Island almost 
parallel to and westward of the Coast Range, although for most 
of this distance these mountains are quite low. 
" Another curious fact which may be mentioned in con- 
nection with the general rainfall of the United States is that, 
generally, the great swampy areas occur in regions of highest 
rainfall. This is true, for instance, of the everglades of 
Florida, where the rainfall is from 50 to 70 inches per year- 
It is also true of the great swampy district lying on the coast 
of North Carolina, where the rainfall is 60 inches per year, 
and upward; also of the swampy district about the mouth of 
the Mississippi River; but is not so true of the celebrated 
swampy district lying to the west of the Mississippi along the 
Gulf coast. 
"It is interesting to notice the effects of the Great Lakes 
on the rainfall visible on most of the maps. In general, it will 
be found that the rainfall is greater on the east shore of Lake 
Michigan than on the west shore. It is to be noted that the 
prevailing winds here reach the lake from the west. Either 
they gather up considerable moisture from the lake which is 
deposited on the east shore, or, what is more probable, the 277 
RIVER AND STREAM WATER 
temperature of the lake is such as to chill the air and cause it 
to deposit more of its moisture on the east shore than on the 
west. Much the same is true of the east shores of Lake Erie 
and Lake Ontario, areas which are small in both cases, because 
the lakes themselves lie east and west. There is, however, a 
distinct increase of rainfall along the southeastern coast of Lake 
Erie and to the east of Lake Ontario. These features can be 
traced on the monthly maps, but more perfectly on the sea- 
sonal ones. The effect seems to be somewhat more marked 
in the cold seasons than in the warm, and it is a noteworthy 
fact that the areas of deep snows in Michigan and New York 
are found to be on the same line. The area of deep snows for 
Southern Michigan is from the middle of the west coast, in the 
vicinity of Manistee, nearly straight across the peninsula; 
the area of deep snowfall in New York is to the eastward of 
Lake Ontario, and,, to some degree, to the southward, in the 
immediate vicinity of the lake. It should also be noted that the 
area for deepest snow in the United States not mountainous 
is along the south shore of Lake Superior, from Marquette 
eastward. This would quite agree with the suggested influence 
of the lakes, in that the air passing over Lake Superior comes 
largely from the northwest, and by the time it reaches the 
coast in question has already received a surcharge of vapor 
chilled by the surface of this lake. 
" Another interesting point is the average rainfall for the 
entire United States. The average of all stations, by States, 
gives for spring 9.2 inches, for summer 10.3 inches, for autumn 
8.3 inches, and for winter 8.6 inches, and a total for the year 
of about 36 inches. It appears that the rainfall over the 
United States generally is quite evenly distributed through the 
year, varying in total amount for the seasons from 10.3 for 
summer to 8.3 for autumn. The spring and summer rainfalls 
are the highest; other things being equal, the rainfalls of spring 
and, next to that, of summer are the most useful for agricul- 
tural operations. 
" With the depth given it is not difficult to get the average 
total rainfall for the entire United States (excluding Alaska, 278 
WATER-SUPPLY 
where we have not sufficient information). For this purpose 
we may take the average for each State and multiply it by the 
area of the State, including water-surfaces. Adding these 
together we get 1407 cubic miles as the average annual total 
of water which descends as rain or snow in the United States. 
The figures for the areas are taken from the census of 1890. 
The annual depth of rainfall which this gives is 29 inches, or 
less than that given by the other method. This is to be ex- 
pected, as the other method gave equal weight to each poli- 
tical division, and these divisions are generally smaller in 
the regions of greater rainfall. 
11 To get some conception of this enormous mass of water 
we may compare it with the contents of the Great Lakes, and 
an approximate comparison is near enough. Lake Ontario is 
about 200 miles long and 70 broad, and its average depth is 
about 40 fathoms. It therefore contains about 636 cubic miles 
of water. The annual rainfall would fill it two times and leave 
something over for a third time. Lake Michigan is about 310 
by 70 miles and has an average depth of about 50 fathoms, 
and consequently contains about 1233 cubic miles of water. 
The average annual rainfall would fill Lake Michigan and leave 
174 cubic miles over. Four years of rainfall would probably 
be enough to fill all the Great Lakes." (U. S. Weather 
Report.) 
The average rainfall of the whole earth has been computed 
at 36 inches per annum.* It is interesting to note that the same 
figure answers for the average for the United States (36.3) 
and also for the average for the State of New York (36.5). 
The wettest place in the world is Cherrapungi, in the Khasi 
Hills, Assam, 4455 feet above sea level, where the annual 
rainfall varied between 500 inches and 600 inches, and in the 
year 1891 it was 920 inches, f 
Middleton J arrives at the available rainfall by taking 
the average, deducting the local evaporation and then deduct- 
* Water and Water Engineering, Nov. 1913. 
t Ibid., paper 313. 
t " Water Supply," p. 52. RIVER AND STREAM WATER 
279 
ing further one-fifth of the average rainfall to allow for three 
successive dry years. 
Binnie considers that a good record covering thirty-five 
years will give an average rainfall that will be correct within 
2 per cent, while Henry holds that forty years are required and 
that even then the result cannot be trusted nearer than 5 per 
cent.* 
" Years of large or maximum rainfall recur at intervals 
of approximately ten years, while the years of low or minimum 
rainfall recur at much longer intervals." f 
SEVERE DROUGHTS IN THE MIDDLE STATES J 
" Mr. C. Warren furnishes the following from records giv- 
ing the length of the most noted dry spells in the Middle 
States: 
In the summer of 1634, 24 days. 
In the summer of 1637^ 74 days. 
In the summer of 1642, 41 days. 
In the summer of 1662, 80 days. 
In the summer of 1664, 45 days. 
In the summer of 1688, 81 days. 
In the summer of 1694, 92 days. 
In the summer of 1705, 30 days. 
In the summer of 1715, 46 days. 
In the summer of 1728, 61 days. 
In the summer of 1730, 92 days. 
In the summer of 1741, 72 days. 
In the summer of 1745, 72 days. 
In the summer of 1754, 108 days. 
In the summer of 1755, 24 days. 
In the summer of 1763, 133 days. 
In the summer of 1773, 80 days. 
In the summer of 1791, 82 days. 
In the summer of 1812, 28 days. 
In the summer of 1856, 26 days. 
In the summer of 1871, 42 days. 
In the summer of 1875, 26 days. 
In the summer of 1876, 26 days. 
" The longest drought above mentioned, which occurred 
in 1763, began on the first day of May, and many inhabitants 
of this country were compelled to send to Europe for grain 
and hay. 
" It should be noted that the longest period of drought in 
the above records occurred over a century ago, at which time 
but little progress had been made in clearing the vast forests, 
draining the ponds, tilling the fields, and making that section 
* J. N. E. Water Works Asso., xviii. 33. 
t Ibid., xviii. 35. 
$ Iowa Weather Service. 280 
WATER-SUPPLY 
of the country habitable for civilized man. A careful study of 
records covering all the years since the early settlement of this 
country does not disclose any appreciable decrease or increase 
in the seasonal precipitation, or in the temperature and hu- 
midity of the air." 
Some Historic Droughts.-" There have been droughts in 
all ages and countries. In the year 310 a.d. hardly a drop of 
water fell in England, and 40,000 people died of famine. 
" The seven years of drought and famine in Egypt, recorded 
in Genesis, began in the year 1708 b.c. 
" In 954 a drought began in Europe, lasting four years. 
The summers were intensely hot and the famine prevailed 
everywhere; 3,000,000 died of hunger. 
"In 1771 an unprecedented drought prevailed through- 
out India. Scarcely any rain fell for a year, and hundreds 
of thousands died of famine, whole districts being depopu- 
lated. 
"In 1837 drought and intensely hot weather prevailed in 
Northwest India. Over 800,000 persons perished from famine. 
Similar destruction was wrought by the same causes in 1865 
and 1866, over 2,000,000 persons perishing of hunger in the 
two years." 
Evaporation measurements, both for land and water sur- 
faces, have been conducted at certain points of the United 
States, and the results have been recorded by such competent 
observers as Desmond FitzGerald, W. J. McAlpine,* Professor 
Fuertes, T. Russell, and others. From the reports of these 
gentlemen and from other official sources the following data 
have been drawn: 
Prof. T. Russell, Jr., obtained the following results from 
* Wm. J. McAlpine, in a report to the Water Committee of Brooklyn in 1852, 
finds " that from 30 to 40 per cent of the falling rain and snow is carried off by 
evaporation." The experiments were made by himself. In the same report the 
quotations are found as to the mean evaporation in inches at the following places- 
Great Britain 32 inches per annum 
Paris 38 " " 281 
RIVER AND STREAM WATER 
experiments designed to show relative rate of evaporation 
as affected by rate of wind motion. Compared with the evap- 
oration from a water-surface when covered by still air, the 
evaporation from the same surface when the wind showed 
different velocities was as follows: Wind 5 miles per hour, 
evaporation 2.2 times. Wind 10 miles per hour, evaporation 
3.8 times. Wind 15 miles per hour, evaporation 4.9 times. 
Wind 20 miles per hour, evaporation 5.7 times. Wind 25 
miles per hour, evaporation 6.1 times. Wind 30 miles per hour, 
evaporation 6.3 times. These experiments were made when the 
temperature of the air was 84° F. and the humidity was 50 
per cent.* 
At Rothamsted, England, with an average annual rainfall 
of 31.04 inches, it was found that evaporation from bare soil 
amounted to 17.09 inches, and that 13.95 inches percolated to 
a depth exceeding five feet, and appeared as drainage. Of 
these 13.95 inches of drainage 9.44 inches collected during 
five months, beginning in October, and the remaining seven 
months furnished only 4.51 inches, showing that the ground- 
water depended upon winter drainage for its principal reinforce- 
ment. f 
" Evaporation from saturated woodland soil is from 61 to 
63 per cent less than from saturated soil in the open, the rain- 
fall in woodland commonly exceedingly the evaporation, even in 
summer. 
" Woldrich found that less water percolated in soil upon 
which grass was growing than upon a bare soil. Very light 
rains were wholly lost by evaporation from the grass. He also 
noted that when the snow melted in the spring the water from 
it passed from it into the bare land quicker and in larger quan- 
tity than it did into the soil that was grass-covered. 
" According to Wollny, a calcareous loam which permitted 
38 per cent of the rainfall to soak through when it was bare 
of vegetation percolated no more than 20 per cent of the rain- 
fall when grass or clover was growing upon it." (Storer.) 
* Iowa Weather and Crop Service, Feb., 1904. 
f J. Chem. Soc., li. 504. 282 
WATER-SUPPLY 
TABLE SHOWING RELATION OF EVAPORATION TO RAINFALL 
(MASSACHUSETTS) 
Month. 
Average Year. 
Year of Low Rainfall (1883). 
Rainfall, 
Inches. 
Evapora- 
tion, 
Inches. 
Excess or 
Deficiency 
of Rainfall, 
Inches. 
Rainfall, 
Inches. 
Evapora- 
tion, 
Inches. 
Excess or 
Deficiency 
of Rainfall, 
Inches. 
January 
4.18 
'0.98 
+3-20 
2.81 
0.98 
+ 1.83 
February 
4.06 
1.01 
+3-05 
3-86 
I .OI 
+ 2.85 
March 
4.58 
i'.45 
-(-3.13 
1.78 
1 -45 
April 
3-32 
2-39 
+°-93 
1.85 
2-39 
-o-54 
May 
3 •20 
3.82 
- 0.62 
4.18 
3 • 82 
+ 0.36 
June 
2.99 
5-34 
-2-35 
2.40 
5-34 
-2.94 
July 
3-78 
6.21 
-2-43 
2.68 
6.21 
-3-53 
August 
4-23 
5-97 
-i-74 
0.74 
5-97 
"5-23 
September 
3-23 
4.86 
-1.63 
1-52 
4.86 
"3-34 
October 
4.41 
3 -47 
+0.94 
5 • 60 
3 -47 
+ 2.13 
November 
4-11 
2.24 
+ 1.87 
1.81 
2.24 
- 0.43 
December 
3-71 
1.38 
+ 2.33 
3-55 
138 
+ 2.17 
45.80 
39.12 
+6.68 
32.78 
39-12 
- 6.34 
Note. + indicates excess of rainfall; - indicates deficiency. 
In the year of low rainfall the evaporation was 6.34 inches 
greater than the rainfall. During the warmer months, from 
April to September inclusive, the excess of evaporation was 
15.22 inches, and during the other six months the rainfall was 
8.88 inches in excess of the evaporation. These figures indi- 
cate that a pond will not lower by evaporation in a dry sum- 
mer more than about fifteen inches, even if it received no 
water from its watershed. 
Just determination of the rate of evaporation is a decidedly 
difficult problem to solve, for there exist so many disturbing 
factors which must be taken into consideration-such as direc- 
tion and force of wind, character of soil, influence of crops, and 
like matters-all of which tend to make the final result one 
of only local application. 
Figures that probably represent the maximum evapora- 
tion rate for the United States are from a year's record at 
Yuma, Ariz. The difference between them and those showing 
the rainfall are marked.* The record was kept by W. D. 
Smith, observer for the U. S. Geol. Survey. 
* Engineering News, 51:248. RIVER AND STREAM WATER 
283 
1903. 
Evaporation, 
Inches. 
Rainfall, 
Inches. 
January 
February 
March 
April 
May 
June 
July 
August 
September 
October 
November 
December 
3-32 
3-48 
5-28 
7.21 
9.84 
10.25 
10.75 
10.70 
905 
5-79 
3.80 
3-39 
.00 
•23 
.00 
.00 
.00 
.00 
.04 
.00 
.67 
.04 
.00 
.00 
Net evaporation 
82.86 
.98 
.98 
81.88 
Plant requirements, although not strictly to be classed 
with evaporation, are commonly thought of in that connection, 
as they represent water losses similar in character. 
" After extensive investigation it has been established that 
the rain which falls upon a crop during its growth is insufficient 
for its maintenance, and that such a crop would die were it cut off 
from drawing upon the reserve water stored up in the ground. 
" About one quarter of a summer rainfall may cling to the 
leaves of trees and evaporate directly therefrom, while at the 
same time the trees act as pumping-engines to dry the ground, 
owing to evaporation from their enormous leaf-surface. Thus 
clay lands often become very wet after the cutting off of the 
trees." (Storer.) 
The following is Risler's table of daily consumption of 
water for different crops, quoted in an article on irrigation 
by W. Tweeddale: * 
Inches. 
Lucern grass, .fromo.134too.267 
Meadow-grass, from 0.122 to 0.287 
Oats fromo.14otoo.193 
Inches. 
Indian corn.f rom o. 11 o to 1.570 
Clover .... from 0.140 to 
* Kansas State Board of Agriculture Report, December 31, 1889. 284 
WATER-SUPPLY 
Inches. 
Vineyard from 0.035 to 0.031 
Wheat from o. 106 to o. 110 
Rye from 0.091 to 
Potatoes from 0.038 to 0.055 
Oak-trees from 0.038 to 0.030 
Fir-trees from 0.020 to 0.043 
Snyder gives similar data in a different form.* He places 
the average amount of water, in tons, required for the produc- 
tion of an average acre of various crops as follows: 
Clover 400 tons of water 
Potatoes 400 
Wheat 350 
Cats 375 
Peas 375 
Corn 300 
Grapes 375 
Sunflowers 6000 
An average rainfall of two inches per month during the 
three months of crop-growth would be equivalent to 369 tons 
of water per acre. 
M. Tweeddale concludes that " from seed-time to harvest 
cereals will take up fifteen inches of water, and grasses thirty- 
seven inches. These conclusions agree with practice in irri- 
gation, and show plainly that the demands of plant-growth 
cannot be ignored in tracing the disappearance of rain. The 
figures also exlpain the low summer flow of streams flowing 
from a highly cultivated watershed. They do not necessarily 
explain the effect of forests in regulating flow, since many 
watersheds, although cleared of trees, are not put under culti- 
vation, but still some change in flow. The action of forests 
is probably largely to retard surface-flow by means of irregular 
surfaces, caused by roots, fallen timber, absorbent mosses, 
and leaf accumulation, thus holding the water until it can 
* Chemistry of Soils and Fertilizers, p. 25. RIVER AND STREAM WATER 
285 
LINES OF EQUAL ANNUAL DEPTH OF EVAPORATION IN INCHES, FROM A FREE WATER-SURFACE. BASED ON OBSERVATIONS 
FROM JULY, 1887. TP TUNE, 1888, INCLUSIVE. 286 
WATER-SUPPLY 
DEPTH OF EVAPORATION, IN INCHES, AT SIGNAL-SERVICE 
STATIONS, IN THERMOMETER SHELTERS 
Computed from the means of the tri-daily determinations of dew-point and wet- 
bulb observations 
Stations and Districts. 
00 
00 
00 
a 
d 
< 
| Feb., 1888. 
00 
co 
CO 
d 
s 
| April, 1888. 
May, 1888. 
June, 1888. 
July, 1887. 
Aug., 1887. 
Sept., 1887. 
Oct., 1887. 
Nov., 1887. 
Dec., 1887. 
Year. 
New England. 
Eastport 
0.9 
1.4 
i-5 
2-4 
2-5 
2-7 
2.2 
2-9 
2-5 
2.6 
2.2 
1-4 
.25.2 
Portland 
1.0 
I . 2 
1.8 
2.6 
1.8 
3-3 
3-8 
3-9 
3-4 
3° 
2-5 
i-4 
29.7 
Manchester 
0.9 
1.6 
2.2 
3-3 
3-8 
5-o 
4-i 
3-3 
2-5 
2.8 
2.4 
1-4 
33-3 
Northfield 
0.8 
1.0 
i-5 
2-3 
2-5 
3-4 
3-5 
2-7 
2-3 
1.8 
1.1 
1.0 
23-9 
Boston 
1.2 
1.6 
2.2 
3-4 
3-i 
4-7 
4-4 
4-0 
3-5 
2-7 
2.2 
1.4 
34-4 
Nantucket 
I. I 
1.1 
I . 2 
i-5 
1.8 
2.1 
3-3 
3-8 
3-4 
2-7 
1.8 
1.8 
25.6 
Wood's Holl 
°-5 
0.8 
1.8 
2-4 
1.8 
2-7 
2-7 
2.4 
2-7 
I . 2 
0.8 
0-5 
20.3 
Block Island 
I. I 
1.1 
I . 2 
2.0 
1.8 
2.6 
2-5 
3-i 
2.8 
2.6 
1.8 
1.4 
24.0 
New Haven 
I. I 
1.6 
1.8 
2-7 
2-7 
4-i 
3-7 
3-8 
3-i 
3-2 
2.4 
1.6 
31-8 
New London 
i-5 
1-3 
i-5 
2.6 
2.8 
4.0 
3-4 
3-9 
3-2 
31 
2.4 
2.1 
31-8 
Mid. Atlantic States. 
Albany 
0.9 
1.2 
1.6 
3-3 
3-9 
4-5 
5-o 
4-7 
3-2 
3-o 
2.1 
i-4 
34-8 
New York City. . . . 
1.8 
1.4 
2.0 
3-4 
3-3 
4-6 
5-o 
5-2 
4-3 
4-i 
3-3 
2.2 
40.6 
Philadelphia 
1.6 
2.1 
2.5 
4-4 
4.0 
5-7 
5-7 
5-2 
4-3 
4-0 
3-3 
2.2 
45-o 
Atlantic City 
1.2 
1.6 
i-5 
2-4 
1.8 
3-6 
2-9 
3-3 
2.4 
1.8 
1.2 
i-5 
25.2 
Baltimore 
2.0 
2.2 
2.8 
5-i 
4-7 
5-6 
6.0 
5-o 
4-4 
4-3 
3-6 
2-4 
48.1 
Washington City. . . 
1.8 
i-7 
2-5 
4.2 
3-8 
6.0 
5-4 
4-9 
4-i 
4-2 
4-5 
2-5 
45-6 
Lynchburg 
2.6 
2.7 
3-4 
3-2 
4-5 
5-6 
4-7 
4-3 
3-3 
3-4 
3-2 
2:6 
45-5 
Norfolk 
1.8 
1.6 
2-3 
3-5 
3-2 
4.2 
4-6 
3-7 
3-7 
2.9 
2-3 
1.8 
35-6 
So. Atlantic States. 
Charlotte 
2.0 
2.6 
4-3 
0.4 
4-5 
5-8 
4.0 
4-o 
4-6 
4-° 
3-6 
2.6 
49-o 
Hatteras 
1.0 
1.6 
1.6 
2-5 
2.2 
3-o 
3-3 
4-1 
3-8 
3-2 
2.0 
1.6 
3i-3 
Raleigh 
2.0 
1.8 
2.6 
3-8 
4-i 
5-4 
4.2 
3-2 
3-o 
2-7 
2.4 
1.8 
37° 
Wilmington 
2.4 
2.2 
2-7 
3-3 
3-3 
4-3 
4-3 
3-i 
3-9 
3-4 
2.8 
2-7 
38.4 
Charleston 
2 . r 
2-5 
3-5 
3-7 
3-9 
4-4 
4-5 
4-8 
4-2 
4-o 
3-2 
2-5 
43-7 
Columbia 
2.2 
2-3 
2.6 
4.8 
4-3 
5-4 
4.2 
3-8 
4-2 
3-4 
3-6 
2-4 
43-2 
Augusta 
3-o 
2.6 
3-4 
5-3 
4-8 
5-o 
4-8 
4-5 
51 
4-i 
3-6 
3-i 
49-3 
Savannah 
3-3 
2.8 
4-i 
4-7 
4-3 
4-6 
4.2 
4-7 
3-4 
3-6 
3-5 
2.8 
46.0 
Jacksonville 
2.9 
2.6 
3-8 
4-3 
4-6 
5-3 
5-o 
4-7 
3-8 
3-6 
3-o 
2.1 
45-7 
Florida Peninsula. 
Titusville 
3-5 
2.6 
3-3 
3-8 
3-8 
4-3 
3-8 
4-3 
4-o 
4-i 
3-6 
3-i 
44-2 
Cedar Keys 
3-3 
2.8 
4.0 
4-6 
4-5 
5-i 
5-o 
5-5 
4-5 
4-i 
3-5 
2.6 
49-5 
Key West 
3-8 
3-7 
3-8 
4-5 
4-4 
4-8 
5-i 
5-i 
4-7 
4-3 
3-8 
3-6 
Si-6 
Eastern Gulf Staf c. 
Atlanta 
2.7 
2.6 
4.0 
6.2 
4-7 
5-o 
4-5 
4-7 
5-8 
4-6 
4-2 
2-5 
Si-S 
Pensacola 
2.9 
2.8 
4-i 
4-o 
4-3 
4-6 
5-o 
5-4 
5-2 
4-5 
3-6 
2-4 
48.8 
Mobile 
2.6 
2-5 
2.8 
3-.5 
3-7 
4-o 
4.1 
4-6 
4-6 
4-i 
3-4 
2.2 
42.1 
Montgomery 
3-5 
3-3 
5-i 
6-5 
5-9 
5-8 
4-3 
4-5 
5-7 
4-6 
4-3 
3-1 
56-6 
Vicksburg 
2 . I 
2-5 
3-6 
5-i 
5-7 
4-8 
4-o 
5-o 
4-7 
3-4 
4-o 
2.2 
47-i 
New Orleans 
2.8 
2.8 
4.1 
3-8 
4.2 
4-i 
4-i 
4-3 
4-4 
4-6 
3-7 
2-5 
45-4 RIVER AND STREAM WATER 
287 
DEPTH OF EVAPORATION, IN INCHES, AT SIGNAL-SERVICE 
STATIONS.-Continued 
Stations and Districts. 
CO 
CO 
00 
d 
d 
I Feb., 1888. 
00 
00 
00 
d 
s 
| April, 1888. 
May, 1888. 
June, 1888. 
July, 1887. 
Aug., 1887. 
Sept., 1887. 
Oct., 1887. 
<X) 
00 
> 
0 
£ 
| Dec., 1887. 
1 Year. 
1 
Western Gulf States. 
Shreveport 
i .6 
2 . I 
3-o 
4.8 
4-9 
4-2 
4-9 
5-2 
5-o 
4-i 
3-4 
2.4 
45-6 
Fort Smith 
2.2 
2.7 
3-5 
5-3 
4-4 
4-6 
5-6 
4-6 
4-7 
5-9 
3-9 
2.2 
49-6 
Little Rock 
2 . I 
2.8 
3-5 
5-5 
4.8 
4-i 
5-4 
5-9 
5-8 
5-2 
4-3 
2-3 
51.7 
Corpus Christi 
1-4 
i.6 
3-3 
3-o 
3-2 
3-9 
4-4 
4-3 
4-3 
4-i 
3° 
2-3 
38.8 
Galveston 
i .6 
2.8 
3-2 
2-9 
4-3 
4-2 
5-3 
5-2 
5-'2 
4-7 
4-2 
2-4 
46.0 
Palestine 
2 . I 
3-o 
3-3 
4.2 
4-3 
4-5 
5-8 
4.6 
4-8 
4-4 
4-o 
2.1 
47-i 
San Antonio 
2.4 
3-3 
4-i 
3-8 
4-0 
4-5 
6.6 
5-8 
5-2 
5-4 
4-2 
3-1 
52.4 
Rio Grande Valley. 
Rio Grande City.. . . 
2.7 
3-5 
3-5 
3-6 
4-5 
4-6 
6.9 
7.0 
5-2 
4-9 
3-6 
3-i 
53-i 
Brownsville 
i.8 
2.6 
2-9 
3-o 
3-5 
3-9 
4-o 
4-i 
3-3 
3-o 
2.6 
2-3 
37-o 
Ohio Vai. & Tenn. 
Chattanooga 
2.0 
3-3 
3-3 
5-3 
3-7 
4.3 
4-3 
5-o 
5-4 
4-0 
3-9 
1.9 
46.4 
Knoxville 
2.4 
2.6 
3-4 
5-o 
3-5 
4-2 
4-9 
5-o 
4-9 
4-i 
3-8 
2 . I 
45-9 
Memphis 
2 . I 
2-3 
3-i 
5-9 
5-3 
4.8 
4-9 
5-4 
5-5 
4-2 
4-i 
2-4 
5°o 
Nashville 
i-9 
2 . I 
3-2 
5-9 
5-o 
5-i 
5-5 
6-3 
5-9 
4-0 
3-3 
1.9 
5o.i 
Louisville 
i-7 
2 . I 
2.8 
5-6 
5-4 
5-8 
6.8 
7-4 
6-4 
4-9 
3-8 
2 . I 
54-8 
Indianapolis 
i-3 
1.4 
2.2 
4.6 
4-8 
5-7 
7-7 
6.9 
5-2 
4-1 
3-i 
1.6 
48.6 
Cincinnati 
i.8 
i.8 
2.6 
4-9 
5-2 
6.4 
6-5 
6.6 
6.1 
4-7 
3-3 
2 . I 
52.0 
Columbus 
i .6 
2.0 
2-3 
4-5 
4-8 
5-8 
6-9 
6.4 
5-i 
4-o 
2.6 
1.8 
47-8 
Pittsburg 
1.4 
1.9 
2.2 
3-8 
4-2 
5-4 
6.6 
5-6 
4-9 
3-4 
2.8 
2-3 
44-5 
Lower Lake Region. 
Buffalo. 
o.8 
i. i 
i-3 
2.2 
3-3 
3-9 
4-9 
5-2 
3-9 
2.8 
1.9 
i .6 
32.9 
Oswego 
o.6 
I .o 
i. i 
2.2 
2.8 
3-8 
3-9 
4-o 
3-6 
2-7 
2.2 
1.0 
28.9 
Rochester 
0.5 
I. I 
°-9 
2.6 
3-8 
4-9 
4-6 
4.1 
3-8 
2.6 
2.2 
1 -3 
32.4 
Erie 
I .o 
1.4 
i-4 
2-7 
3-7 
4-6 
5-5 
4.8 
3-i 
2-5 
1.9 
I . 2 
33-8 
Cleveland 
i. i 
1-4 
i-5 
2.9 
3-3 
4-4 
5-2 
4-9 
3-8 
3-4 
2-4 
1.4 
35-7 
Sandusky 
o.8 
1.4 
i-5 
3-2 
3-7 
4.6 
5-4 
5-4 
3-7 
3-4 
2.2 
i-3 
36.6 
Toledo 
°-9 
i. i 
i-5 
3-5 
3-8 
4-6 
6.0 
6.4 
3-7 
3-4 
2.4 
1-3 
386 
Detroit 
o.8 
I. I 
i. 6 
3-0 
4-i 
4-8 
5-9 
5-2 
3-4 
2.8 
2.0 
i-3 
36.0 
Upper Lake Region. 
Alpena 
0.7 
o.6 
o-9 
1.6 
2.1 
3-6 
3-8 
3-7 
2.8 
2.2 
i-5 
0.8 
24-3 
Grand Haven 
o-5 
0.7 
i-3 
2.6 
3-i 
3-8 
4-7 
3-8 
2-7 
2.6 
i-7 
1.1 
28.6 
Lansing 
o.6 
I . 2 
i-4 
2-7 
2.8 
4-o 
4-3 
3-9 
2.4 
1.9 
i-4 
1.0 
27.6 
Marquette 
o.8 
o.8 
0.9 
i-7 
2-4 
3-3 
3-4 
3-3 
3-i 
2.2 
i-3 
i-3 
24-5 
Port Huron 
o.6 
I .o 
I. I 
2.6 
3-o 
3-8 
4.6 
4-2 
3-2 
2-5 
i-7 
1.0 
29-3 
Chicago 
I .o 
I . 2 
1.8 
3-2 
3-3 
4-8 
5-4 
5-3 
4-i 
3-2 
2-3 
1.2 
36.8 
Milwaukee 
°-5 
I .O 
I. I 
2.4 
2.6 
3-8 
4-8 
3-7 
3-4 
2.9 
1.9 
0.9 
29.0 
Green Bay 
o-5 
o.6 
0.8 
i-7 
2-5 
4-i 
5-6 
4-2 
3-o 
2.4 
1.9 
0.9 
28.2 
Duluth 
o-5 
o-5 
0.6 
i-5 
2.4 
2-5 
3-9 
3-4 
3-0 
2.5 
1.2 
1.0 
23-0 
Extreme Northwest. 
Moorhead 
O. 2 
i-4 
°-5 
2.1 
3-6 
3-8 
3-7 
3-3 
3-5 
2.4 
i-3 
o-5 
26.3 
Saint Vincent 
0-3 
o-3 
o-5 
1.8 
3-8 
3-9 
3-i 
2.6 
2.6 
2.0 
0.9 
o-3 
22.1 288 
WATER-SUPPLY 
DEPTH OF EVAPORATION, IN INCHES, AT SIGNAL-SERVICE 
STATIONS-Continued 
Stations and Districts. 
00 
00 
00 
d 
oj 
Feb., 1888. 
Mar., 1888. 
| April, 1888. 
May, 1888. 
June, 1888. 
July, 1887. 
Aug., 1887. 
Sept., 1887. 
i) 
00 
0 
O 
Nov., 1887. 
Dec., 1887. 
Year. 
Extreme Northwest. 
Bismarck 
0-4 
o.6 
o.6 
3-o 
4-3 
4-1 
5-6 
4.2 
4-0 
2.6 
1.2 
0.4 
31-0 
Fort Buford 
i-4 
0.7 
o.6 
3-o 
4-7 
5-o 
6.2 
49 
4-8 
30 
i-7 
0-5 
35-5 
Fort Totten 
0.2 
o-3 
0.4 
2.2 
4.6 
3-8 
4-2 
3-7 
3-7 
2-3 
1.4 
0.4 
27.2 
Up. Mississippi Vai. 
Saint Paul 
0.7 
0.7 
2.2 
2.0 
2-3 
4-i 
5-o 
3-7 
2.8 
2-4 
i-5 
0.7 
28.1 
La Crosse 
o-4 
I . 2 
i-4 
3-3 
3-5 
4-4 
5-4 
4-7 
3-o 
30 
1.8 
0.8 
32.9 
Davenport 
0-5 
I .O 
i.8 
3-8 
3-4 
4.6 
6.9 
6.2 
4-4 
30 
2-3 
1.1 
39-0 
Des Moines 
o.6 
I .o 
i-5 
3-7 
3-i 
4-2 
6.6 
4-7 
4.1 
3-3 
2-3 
0.9 
36.0 
Dubuque. 
o-7 
I .o 
i-4 
2.2 
2.9 
4-2 
6.2 
4-8 
3-3 
2.8 
1.8 
0.9 
33-2 
Keokuk 
o.8 
I. I 
2.1 
4-2 
3-7 
4-3 
7-o 
6.8 
5-o 
3-8 
2.9 
1.2 
42.9 
Cairo 
i .6 
2 . I 
2.9 
5-8 
4-4 
4-3 
5-6 
6-5 
5-i 
4-5 
3-8 
2-3 
48.9 
Springfield, III 
o.8 
I . I 
2 .O 
4-6 
3-8 
4-3 
5-4 
6-5 
4-5 
3-5 
2.9 
1.4 
40.8 
Saint Louis 
i-3 
i .6 
2-5 
5-5 
4-7 
5-o 
7-5 
8.0 
5-9 
4-9 
3-9 
i-4 
52.2 
Missouri Valley. 
Lamar 
I. I 
i .6 
2-4 
4-4 
3-8 
4-o 
6.0 
4.6 
3-7 
3-6 
2.9 
i-5 
39-6 
Springfield, Mo 
I. I 
i-7 
2-4 
5-o 
4-8 
4-o 
5-o 
3-4 
3-4 
3-5 
3-i 
i-4 
38.3 
Leavenworth. 
o-9 
i-5 
2:3 
4.6 
4-5 
5-o 
6-3 
4-5 
4-0 
3-9 
2-7 
1.4 
41.6 
Topeka 
I. I 
I . 2 
2.0 
4-o 
4.1 
4-i 
6-3 
3-5 
3-2 
3-o 
2.2 
i-4 
36 1 
Omaha 
o.8 
i-5 
1-4 
4-4 
3-8 
5-2 
6.2 
5-2 
4-3 
4-3 
3-o 
1-4 
4i-7 
Crete 
0.7 
I. I 
1.2 
3-5 
3-3 
4-5 
5-6 
4-7 
3-8 
3-6 
2-4 
1.1 
35-5 
Valentine 
I . 2 
i .6 
1.8 
5-o 
3-2 
5-3 
6.9 
5-o 
5-2 
3-8 
3-3 
i-5 
43-8 
Fort Sully 
o.6 
o-9 
i-3 
4-4 
4-i 
5-2 
7-7 
4-9 
5-7 
3-6 
2.8 
0.7 
41.9 
Huron 
o-3 
0.7 
0.8 
3-7 
3-7 
4-i 
5-7 
4-2 
4-i 
3-i 
2-4 
0.7 
33-o 
Yankton 
0-4 
i-4 
1.2 
3-3 
3-i 
4-4 
4-6 
3-7 
2.9 
3-o 
2.2 
0.8 
31.0 
Northern Slope. 
Fort Assiniboine.. .. 
o.8 
I . 2 
1.2 
3-8 
4-i 
4-2 
6.8 
5-5 
4-8 
3-5 
2-5 
1.1 
39 5 
Fort Custer 
o.6 
1-5 
i-3 
5-4 
6.8 
4-9 
9.6 
8.0 
6.1 
3-4 
2.9 
i-5 
52.0 
Fort Maginnis 
I. I 
i-4 
1.1 
3-3 
3-2 
4.6 
6.8 
4.6 
3-8 
2.8 
2.0 
1.1 
35-8 
Helena 
I. I 
3-6 
2.1 
6.1 
4-3 
5-5 
7-2 
7-7 
6.4 
4-3 
3-0 
2.1 
53-4 
Poplar River 
o-4 
o.8 
0.8 
2-7 
4-9 
5-7 
6.0 
4-8 
4-4 
2-5 
1-7 
o-7 
35-4 
Cheyenne 
3-3 
5-7 
4-0 
8.2 
5-2 
10.4 
8.0 
7-7 
8.6 
5-8 
6.1 
3-5 
76-5 
North Platte 
o.8 
i.8 
1.8 
5-4 
3-9 
6.9 
6.0 
4-8 
3-7 
2.8 
2-3 
1.1 
4i-3 
Middle Slope. 
Colorado Springs. . . 
3° 
3-3 
4-i 
6-7 
5-6 
4-3 
6-7 
7-2 
6.8 
4-6 
4-2 
2.9 
59-4 
Denver 
2.8 
3-7 
3-5 
7-6 
5-8 
10.5 
8-3 
8-5 
6.1 
4-9 
4.2 
3-i 
69.0 
Pike's Peak 
2 . I 
i-3 
i-5 
2.1 
1.8 
i-9 
3-o 
4-o 
3-o 
2-3 
2.8 
1.0 
26.8 
Concordia 
i-3 
2.8 
1.8 
4.8 
4-3 
5-7 
7-3 
5-2 
4-3 
4-5 
3-4 
1.8 
47-2 
Dodge City 
i-4 
2.4 
2.8 
4-i 
4.6 
7-4 
8-3 
6.6 
5-5 
5-2 
4-2 
2.1 
54-6 
Fort Elliott 
i-3 
i-9 
3-2 
5-i 
5-4 
8.2 
7-6 
6.2 
5-4 
4-7 
4-2 
2.2 
55-4 
Southern Slope. 
Fort Sill 
i .6 
2 .O 
2.6 
3-8 
4-0 
4-4 
4-8 
7-5 
5-i 
4-2 
4-i 
2.0 
46.1 
Abilene 
i.8 
i-7 
3-i 
4-2 
5-o 
5-8 
95 
7-5 
6.2 
4-5 
3-4 
i-7 
54-4 RIVER AND STREAM WATER 
289 
DEPTH OF EVAPORATION, IN INCHES, AT SIGNAL-SERVICE 
STATIONS-Continued 
Stations and Districts. 
Jan., 1888. 
Feb., 1888. 
Mar., 1888. 
00 
00 
00 
& 
< 
May. 1888. 
June, 1888. 
July, 1887. 
Aug., 1887. 
Sept., 1887. 
Oct., 1887. 
| Nov., 1887. 
| Dec., 1887. 
Year. 
Southern Slope. 
Fort Davis 
5-4 
5-7 
6.7 
8.5 
II .0 
12 .O 
11.4 
9.0 
5-9 
5-2 
5-7 
4.9 
96-4 
Fort Stanton 
3-9 
3-9 
5-2 
7-3 
9-5 
10.9 
9-4 
11.6 
3-9 
4-o 
3-6 
3-8 
76.0 
Southern Plateau. 
El Paso 
4.0 
3-9 
6.0 
8.4 
10.7 
13.6 
9-4 
7-7 
5-6 
5-2 
4.6 
2.9 
82.0 
Santa Fe 
3° 
3-4 
4-2 
6.8 
8.8 
12.9 
9.2 
9.8 
6.6 
6.7 
5-7 
2-7 
79.8 
Fort Apache 
2.6 
3-o 
3-6 
6.8 
94 
9.1 
7-i 
6.7 
5-3 
5-2 
41 
2.6 
65-5 
Fort Grant 
5-2 
4-8 
6-4 
9-2 
10.2 
13-8 
12.4 
io-5 
9.0 
7-9 
7-2 
4.6 
101.2 
Prescott 
i-4 
2.8 
3-6 
5-4 
6.2 
8.1 
6.6 
6-5 
4-7 
4-9 
3-6 
2.2 
56.0 
Yuma 
4-4 
5-2 
6.6 
9.6 
9.6 
12.6 
II .0 
10.2 
8.2 
8.2 
5-5 
4.6 
95-7 
Keeler 
3° 
4.6 
6-3 
8-7 
9-3 
11.9 
12.8 
i3-9 
10.6 
8.8 
5-9 
4-8 
100.6 
Middle Plateau. 
Fort Bidwell 
0.8 
1.8 
1.8 
4.6 
5-2 
4-o 
8.8 
8.1 
5-o 
4.6 
2.4 
i-3 
48.9 
Winnemucca 
0.9 
2.8 
6.2 
9-i 
9-3 
IO. I 
n-5 
12.0 
9.9 
6.6 
3-7 
1.8 
839 
Salt Lake City 
1.8 
2-7 
3-6 
7-2 
6.9 
8.9 
9.2 
10.7 
9.6 
6-5 
5-o 
2-3 
74-4 
Montrose 
1.8 
2-7 
3-7 
6.2 
7.0 
II. I 
IO. 2 
8.3 
6.9 
5-2 
3-4 
2.0 
68.3 
Fort Bridger 
Northern Plateau. 
1.6 
2-5 
2-7 
4-3 
4-3 
6-5 
7-7 
6.8 
5-6 
4-2 
5-2 
4-7 
56.1 
Boise City 
1.6 
2-5 
3-8 
6.1 
6-5 
6.6 
10.0 
9-2 
7-4 
5-2 
3-2 
1.8 
63-9 
Spokane Falls 
0.7 
i-7 
2-7 
4-4 
5-4 
4-4 
7-7 
6.4 
3-8 
2.5 
i-7 
1.4 
42.8 
Walla Walla 
I. I 
2.9 
3-6 
6.2 
7-7 
5-7 
9-9 
7-9 
5-i 
3-4 
1.8 
2.4 
57-7 
North Pacific Coast. 
Fort Canby 
I . 2 
I. I 
1.8 
2.1 
2.8 
2-3 
1.8 
2.9 
1.8 
1.8 
i-5 
0.9 
21.1 
Olympia 
i-3 
I . 2 
1.8 
2-5 
4-i 
3-3 
3-2 
3-i 
2-4 
i-5 
i-3 
I. I 
26.8 
Port Angeles 
1.0 
0-9 
1.8 
1.8 
2-5 
2 . I 
2.1 
1.8 
i-5 
1.2 
i-3 
I. I 
19.1 
Tatoosh Island 
I . 2 
1.1 
1.8 
i-4 
1.8 
1.8 
1-4 
i-4 
1-4 
1.6 
1.8 
i-4 
18.1 
Astoria 
I. I 
1.0 
1.6 
2 . I 
3-o 
2-7 
3-o 
2-9 
2.6 
2-3 
1.8 
1.2 
25-3 
Portland 
0.9 
I. I 
2.4 
3-4 
5-o 
3-2 
5-4 
4-2 
3-4 
2-7 
1.8 
I . 2 
34-7 
Roseburg 
I . 2 
1.6 
2-7 
3-9 
4-7 
3-5 
5-4 
4-7 
5-o 
3-2 
1-7 
1.6 
39-2 
Middle Pacific Coast. 
Red Bluff 
3-0 
4-6 
5-4 
6.1 
7-o 
6.9 
II .0 
10.7 
10.1 
io-5 
5-9 
3-6 
84.8 
Sacramento 
1.8 
3-i 
3-7 
4-3 
4-2 
5-6 
5-9 
5<5 
6-5 
7-3 
3-9 
2-4 
54-3 
San Francisco 
2-7 
2-7 
3-3 
3-i 
2.8 
3-i 
2-4 
2-5 
3-3 
5-o 
2.8 
3-o 
36.7 
South Pacific Coast. 
Fresno 
1.8 
2.8 
3-o 
5-6 
6.0 
7.0 
9.1 
10.2 
7.6 
6-7 
3-8 
2.2 
65.8 
Los Angeles 
2-3 
2.0 
2.8 
3-4 
3-o 
3-8 
3-2 
3-5 
3-i 
4-i 
3-0 
3-o 
37-2 
San Diego 
2.9 
2-7 
2-5 
2-7 
3-3 
2.8 
3-2 
3-3 
2.9 
4-3 
3-2 
3-7 
37-5 
be taken into the ground. This is not mere theory; it is based 
on observations made during many days spent in the forest, 
and is believed to almost, if not fully, account for the better 
sustained flow of forest streams and their lighter flood-flows." 290 
WATER-SUPPLY 
The official chart and figures found on pages 285 to 289 
showing depths of evaporation are from the United States 
Weather Service Report: 
Evaporation takes place from both ice and snow. Its 
amount is appreciable even below io° F. and it increases rapidly 
with rise of temperature. 
This fact is shown graphically by A. F. Meyer in his compre- 
hensive paper on 11 Computing Run-off from Rainfall and other 
data." * 
The flow of streams depends upon causes quite various in 
character, such as deep-seated springs, melting of glaciers 
(e.g., the River Rhone), and other unusual sources, but for 
the great majority of cases the flow is traceable directly to the 
rainfall run-off and to springs of local origin. 
In Kicking Horse Valley, British Columbia-, the streams 
shrink on rainy days and swell on those of sunshine for the 
reason that rainfall does not compensate for the lack of the 
melting snow. 
What is the amount of the run-off to be expected per square 
mile of watershed is essentially a local question, not one to be 
formulated generally for widely separated localities. 
Fanning believes that for ordinary watersheds it is fair to 
assume that 50 per cent of the annual rainfall flows off in the 
streams. 
More in detail, it is as follows: 
Mountain slope or steep rocky hills 80 to 90 per cent 
Wooded swampy lands 60 to 80 " 
Undulating pasture and woodland. . . .- 50 to 70 " 
Flat cultivated lands and prairie. 45 to 60 " 
He also considers the " low rain-cycles " mean rainfall to 
be about 80 per cent of the general mean rainfall. 
From the report of the committee on yield of drainage areas, 
J. N. E. Water-works Asso., Dec. 1914, we note the following 
comparison of run-off, upon sundry eastern watersheds, in years 
of high and low precipitation: 
* Am. Soc. C. E., 79:1079. RIVER AND STREAM WATER 
291 
Average 
Average 
Per cent 
Precipitation 
Run-off 
of 
(Inches) 
(Inches) 
Run-off 
Highest 20 years 
5^-29 
30.00 
53-3 
Lowest 20 years 
39-29 
16.92 
43-i 
It is to be observed that when the precipitation is high 
the percentage of it that passes into the run-off is much higher 
than that which follows a smaller amount of rainfall. 
An examination of the data secured by the same committee 
permits of the conclusion that, as a fair average of the Eastern 
watersheds considered by them the land surface will yield for 
storage about one million gallons per square mile per day. 
According to the U. S. government reports, " for the area 
of the United States east of the ninety-fifth meridian the run-off 
is from 35 to 50 per cent of the total rainfall. It appears to 
be largest in the vicinity of the Great Lakes, and diminishes 
from this region slowly to south and east, and rapidly towards 
the west. In the lower peninsula of Michigan, for instance, the 
run-off is 50 per cent of the total rainfall. Along the Gulf coast 
it appears to be only from 30 to 40 per cent, and along the Atlan- 
tic coast it probably varies from 30 to about 50 per cent. In 
general, for the interior States east of the ninety-fifth meridian 
the run-off is between 40 and 50 per cent of the total rainfall. 
" As soon as we cross the ninety-fifth meridian westward 
we find a very sharp fall in the percentage of run-off to the 
total rainfall. For the band extending north and south be- 
tween the ninety-fifth and one hundred and fifth meridians 
this percentage varies from 10 to 25 per cent, and over Iowa 
is about 33 per cent. The percentage is highest at the northern 
end of the band indicated, and lowest at the southern end. 
Going still farther westward we come to another very marked 
area, that of the Continental Divide; here the percentage 
of run-off suddenly increases, reaching the highest figure to be 
found in the United States. From Montana to Colorado 
it varies from 60 to 70 per cent of the total rainfall. In New 
Mexico it falls to about 33 per cent. This evidently on account 
of the easy flow of water from the mountain ranges in the 292 
WATER-SUPPLY 
area in question. West of the Divide the run-off is again small, 
being only 15 or 20 per cent in Arizona and Nevada, about 
30 per cent in Idaho, and nearly 50 per cent in Utah. Utah, 
it seems from its topography, partakes of the character of the 
band lying just to the east of it. Along the Pacific coast the 
run-off is about 25 per cent in Oregon, 30 per cent in Wash- 
ington, and between 45 and 50 per cent in California. 
" In general we may say that the run-off on the more level 
areas of the United States is less than 50 per cent, and on the 
great plains may fall as low as 10 per cent. In the mountain 
regions it may rise to as high as 70 per cent. In the relatively 
dry area, or the areas of distinctly dry seasons, the percentage 
is very much reduced." * The " run-off " for the State of 
Connecticut is placed at 60 per cent of the rainfall. 
The watershed of the new Scituate reservoir to supply 
Providence, R. I., has a rainfall of forty-five inches annually 
and a run-off of twenty-one inches, or 47 per cent. Since 
twenty-one inches of yearly rainfall practically equals one mil- 
lion gallons per square mile per day, we note the probable daily 
yield for storage per square mile to be about one million gallons. 
The following figures have been taken from official sources: 
RAINFALL AND RIVER-FLOW FOR THE CONNECTICUT RIVER 
BASIN 
Month. 
Average Rain 
in Inches. 
River-flow in Inches 
On Whole Watershed. 
Per Cent of River- 
flow to Rainfall. 
January 
3-27 
i-93 
59 1 
February 
3.10 
2.04 
65.8 
March 
3-94 
3.00 
76.3 
April 
3-26 
4-73 
I45-O 
May 
3i7 
4-i9 
132.2 
June 
4.00 
1.46 
36.5 
July 
4-79 
1.02 
21.3 
August 
4-87 
1.06 
21.8 
September 
3-°4 
0.89 
29-3 
October 
3-93 
1.11 
28.3 
November 
3-93 
1.76 
44.8 
December 
3-39 
2.06 
60.7 
44 69 
25-25 
56.5 
* " The point at which a region may be classed as arid and unfit for success- 
ful agriculture without irrigation should be lowered, it is believed, to 15 inches 
annual rainfall." (Report of Chief Signal Officer, 1880.) RIVER AND STREAM WATER 
293 
RAINFALL AND RIVER-FLOW FOR THE POTOMAC RIVER 
BASIN 
Month. 
Rainfall in Inches. 
River-flow in Inches. 
Per Cent of River 
flow to Rainfall. 
January 
3.21 
2.09 
65.2 
February 
3-35 
336 
March 
4-39 
3.62 
82.6 
April 
348 
3-51 
May 
5H 
2.36 
46.3 
June 
5-25 
i-93 
36.8 
July 
4-89 
20.5 
August 
3.81 
.0.78 
20.5 
September 
3-86 
1.06 
27-5 
October 
2.65 
1.21 
45-7 
November 
2 88 
i-79 
1-32 
62.3 
5i-i 
December 
2-59 
45-47 
24 03 
53-o 
Rafter in " Water Supply and Irrigation Papers No. 80, 
U. S. Geological Survey," gives yearly average data for twelve 
drainage basins in the eastern part of the United States, as 
follows: 
Drainage Basins. 
Years of 
Rainfall, 
Run-off, 
Evapora- 
Record. 
Inches. 
Inches. 
tion. Inches. 
i. Muskingum River, Ohio. . .. 
1888-1895 
39-7 
13 I 
26.6 
2. Genesee River, N. Y 
1890-1898 
40.3 
14.2 
26.1 
3. Croton River, N. Y 
1877-1800 
49.4 
22.8 
26.6 
4. Lake Cochituate, Mass 
1863-1900 
47-i 
20-3 
26.8 
5. Sunbury River, Mass 
1875-1900 
46.1 
22.6 
23-5 
6. Mystic Lake, Mass 
1878-1895 
44-i 
20.0 
24 • 1 
7. Neshaminy Creek, Pa 
1884-1899 
47-6 
23-1 
24-5 
8. Perkiomen Creek, Pa 
1884-1899 
48.0 
23 • 6 
24.4 
9. Tohickon Creek, Pa 
1884-1898 
5° •1 
28.4 
21.7 
10. Hudson River, N. Y 
1888-1901 
44-2 
23-3 
20.9 
11. Pequannock River, Conn. . . 
1891-1899 
44-2 
26.8 
20.0 
12. Connecticut River, Conn. ... 
1872-1885 
43-o 
22.0 
21 ,O 
It must be added, however, that Mr. Rafter insists upon 
the danger of depending too much upon " average " results 
in such matters. He says, " What is wanted is a clear state- 294 
WATER-SUPPLY 
ment of the minimum, together with the longest period which 
such minimum may be expected to occupy." 
As a matter of fact all run-off statistics should be used with 
much care. Mr. T. U. Taylor has well said: 
" Any attempt to apply run-off data obtained in one section 
of the country to a locality or stream in another section is full 
of the utmost danger. The regularity of flow, the quantity 
of daily flow, the mean monthly flow, the mean annual flow, 
and the run-off are data that can be obtained only by care- 
ful observations and measurements for the locality, station, 
and stream concerned. The run-off for the same stream, 
at different stations along its course, will be found to vary 
greatly. 
" Run-off is a complex factor, depending on topography, 
vegetation, kinds of soil (whether cultivated or uncultivated), 
rainfall, distribution of rainfall as to time of year and as to 
growth of vegetation, and, what is still more vital, the condition 
of the soil at the time of the rains. Personal observation has 
convinced me that a two-inch rainfall in twenty-four hours 
may at one time give a run-off of 25 per cent and at another 
time no run-off whatever." * 
A curious instance of a most unbalanced relationship 
between rainfall and run-off was reported some years ago.f 
It seems that in the Painted Desert, Ariz., a copious shower 
of rain was observed to fall toward the earth, but it became 
entirely evaporated as it fell and never reached the surface 
of the ground. 
Where conditions are exceptionally favorable for percola- 
tion none of the run-off may reach living streams. 
G. Lidy J uses the following equation: 
P+E+R=R', 
where P is the percolating water entering the soil; E the water 
* Am. Soc. C.E., 40: 166. 
f J. Fk. Inst., Apr., 1904. 
$ " La Technique Sanitaire," p. 145, 1910 RIVER AND STREAM WATER 
295 
which evaporates directly from the surface; R the run-off, 
and R' the rainfall. 
The latest German value obtained for P+R is .4 to .5 
of the value of R', the value of E being at least .8 of R'. These 
values leave the above equation unbalanced, the first member 
being larger than the second. Lidy considers that the dif- 
ference between the two members is caused by an increased 
value of P due to atmospheric moisture condensed directly 
in the pores of the soil from the air which circulates therein. 
Moisture from such a source is, of course, not a portion of the 
rainfall. That it may amount to something noteworthy in 
volume is shown by Lidy's experiments. 
A good map showing the " run-off " for the United States 
will be found in the 14th report of the U. S. Geological Survey, 
Part II, page 150. 
Even with uniformity in rainfall the rate of river-flow must 
vary, owing to such disturbing factors as frozen ground in 
winter and excessive evaporation in summer. . 
For Eastern Massachusetts Mr. Desmond FitzGerald places 
the months in order of dryness, averaging as follows: 
1. July. 
2. September 
3. August. 
4. June. 
5. October. 
6. November. 
7. May. 
8. December. 
9. January. 
io. April. 
ii. February. 
12. March. 
Showing the wettest months to be the first four in the year. 
INFLUENCE OF FOREST UPON WATER SUPPLY 
Do forests change climate; do they increase rainfall; do 
they have any effect upon floods; do they conserve water sup- 
ply? 
Many writers, official and otherwise, have undertaken 
to answer these questions and have done so in fashions so 
various that the ordinary man is forced to fall back upon his 
own resources and attempt the answers for himself. 
A part of the following is freely condensed and extracted 
from the government report upon " Forest Influences," issued 
by B. E. Fernow, Chief of the Forestry Division. 296 
WATER-SUPPLY 
Transpiration occurs through the leaves of trees as it does 
through those of smaller vegetable growths and a considerable 
part of the water falling as rain upon a forest is returned to the 
atmosphere by such action. 
During the period of vegetation the following varieties 
transpire per pound dry weight of leaves: 
Pounds of Water 
Birch and linden 600-700 
Ash 500-600 
Beech 450-500 
Maple 400-450 
Oak 200-300 
Spruce and Scotch pine 50-70 
Fir 30-40 
Black pine 30-40 
Conifers transpire one-sixth to one-tenth of the amount 
which is needed by deciduous trees. 
The transpiration from leaves in full sunshine is decidedly 
greater than from leaves in the diffused daylight or darkness. 
The absolute amount of annual transpiration as observed in 
forests of mature oaks and beeches in Central Europe is about 
one quarter of the total annual precipitation. 
Raphael Zon * of the U. S. Forest service quotes data 
obtained from the Mariabrunn (Austria) forest experiment 
station which show that " one acre of oak forest, 115 years 
old, absorbed in one day from 2227 to 2672 gallons of water per 
acre, which corresponds to a rainfall of from 0.09 to 0.115 
inch per day, or 2.9 to 3.9 inches per month. Taking the period 
of vegetation as five months, the absorption of water would be 
158,895 cubic feet, which represents a rainfall for this period 
of 17.7 inches. This amount of water is given off merely 
through transpiration from the leaves and does not include 
the physical evaporation from the surface of twigs, branches, 
and leaves. These figures, while only approximate, give an 
idea of the enormous quantities of water given off by forests 
* Science, July 18, 1913. RIVER AND STREAM WATER 
297 
into the air. It may be compared to clouds of exhaust steam 
thrown into the atmosphere." 
Evaporation from the ground in the forest is naturally much 
less than in the open fields. 
The forest cover, and especially the litter of a well-kept 
forest, may decrease the amount of evaporation within the 
forest to nearly seven-eighths of that in the open. The reason 
for this important influence of the forest is due not only to 
the impeded air circulation, but also the temperature and 
moisture conditions of the forest air and forest soil. 
It is this protection against evaporation which gives to the 
forest one of its chief values as a guardian of water supply. The 
forest floor, with its irregularities and its sponge-like qualities, 
moreover, stops the rapid and ruinous draining of the surface, 
with possible denuding of the land and favors slow percolation 
through the soil and reinforcement of the springs. 
The influence of forest cover upon the flow of springs is due 
to reduced evaporation, as well as to the fact that by the pro- 
tecting cover the soil is kept granular and allows more water 
to penetrate and percolate than would otherwise be the case. 
The condition of the forest floor is of great importance. Where 
the litter and humus mold is burned up, as in many of our 
mountain forests, this favorable influence is largely destroyed 
although the trees are still standing. 
In consequence of deforestation evaporation from the soil 
is augmented and accelerated, resulting in unfavorable condi- 
tions of soil humidity and affecting unfavorably the size and 
continuity of springs. 
Inasmuch as a low growth of " saplings " may cover a former 
forest floor very quickly, it is difficult to be sure that recent 
data, supposed to be furnished by 11 de-forested " areas, truly 
represent the relation between forest and stream flow on the 
one hand and cleared land and stream flow on the other. 
Snow is held longer in the forest and its melting is retarded, 
giving longer time for filtration into the ground, which also, 298 
WATER-SUPPLY 
being frozen to lesser depth, is more apt to be open for sub- 
terranean drainage. Altogether forest conditions favor, in 
general, greater subterranean and less surface drainage, yet 
the moss or litter of the forest floor retains a large part of the 
precipitation and prevents its filtration to the soil, and thus 
may even diminish the supply to springs. This is especially 
possible with small precipitations. 
Although the quantity of water offered for drainage on 
naked soil is larger, and although much is utilized by the trees 
in the process of growth, yet the influence of the soil-cover in 
retarding evaporation and run-off tends to offset this loss, as the 
soil-cover is not itself dried out. 
Naturally, a saturated forest floor can absorb no further 
amount of water, therefore the run-off from excessive rainfall, 
or rapid melting of snow, must pass to the streams as quickly 
as in a cleared district, but the temporary retention of a large 
amount of water and its eventual partial deflection into sub- 
terranean drainage, with consequent lengthening in the time of 
flow, would at least tend to reduce the number and height of 
regular floods. 
Floods due to special and unusual causes, such as that of the 
upper Hudson River Valley in March, 1913, cannot be guarded 
against. Upon that occasion excessive, long-continued rain 
fell upon a deeply frozen soil devoid of snow. The run-off 
was, of course, like that from a roof. 
The New York Forest Commission, speaking of floods in 
the Adirondack region and the influence of forests in relation 
to them, says: 
" In the uplands of the preserve there are many densely 
wooded tracts adjacent to others from which the forests have 
been stripped. The residents agree that in the former floods 
are unknown, while in the latter they are a yearly occurrence. 
Their appearance was coincident with the disappearance of the 
woods. It was then noticed that the bridges, which for many 
years had sufficed to span the streams during heavy rains, were 
no longer safe, and new ones with. longer spans became a 
necessity." RIVER AND STREAM WATER 
299 
They refer also to the effect of the removal of the forests 
in the Adirondack watersheds upon the navigation of the canals 
of the State and the whole system of inland commerce. 
Placing reliance upon the testimony of " the oldest inhabi- 
tant " involves considerable risk, and is often met by such 
records as the following.* Translation by W. L. Moore. 
Dates of the Inundations of the Seine 
Valley. 
Height at the Bridge 
of La Tournelle. 
Feet. 
Mean per Half 
Century. 
Feet. 
July ii, 1615 
29-99J 
January, 1649 
25-10 J 
27-53 
January, 1651 
25-59 1 
March 1, 1658 
28.87 I 
26.36 
March, 1690 
24.61 J 
March, 1711 
24.77 ] 
December 25, 1740 
25-92 J 
25-34 
January, 1751 
21.98 1 
November 14, 1764 
22.97 
March 4, 1784 
21.85 [ 
22.42 
February 4, 1799 
22.87. 
January 3, 1802 . 
24 • 44 
March 3, 1807 
21.85 
May, 1836 
18.66 
21.22 
February, 1850 
19.91 J 
11 The deductions from this table are striking. The con- 
tinued decrease of the floods for each half century is remark- 
able. The waters attained a mean height of 27.53 feet in the 
first half of the seventeenth century; they only attained a 
mean of 21.22 feet in the present. According to this, we have 
experienced an amelioration of nearly 6.56 feet, and yet the trees 
have been steadily and unceasingly cut down, and the forests 
transformed into cultivated farms." 
Before being convinced by the above we must however 
remember that the most disastrous flood that the valley of the 
Seine ever experienced occurred in 1910. 
As to actual aid in the production of rain-fall it must be 
admitted that it is hard to gather together numerical facts 
that show the forest to be of material value; but as a scientific 
* Annales des Ponts et Chaussees, 1852, p. 102. 300 
WATER-SUPPLY 
proposition the following point raised by Dr. Raphael Zon * 
is worthy of careful attention. 
" While direct evaporation from the ground not sheltered 
by forest-cover may become greater, yet the more rapid run-off 
and the absence of transpiration by trees would necessarily 
reduce the total amount of water evaporated into the atmos- 
phere. The land, were it even taken up for agriculture, would 
not return such large quantities of rain into the atmosphere 
as the forests did. The inevitable result would be that less 
moisture would be carried by the prevailing winds into the 
interior of the country, and therefore less precipitation would 
occur there. The destruction of forests, especially if it leaves 
the ground bare or partly covered with only weak vegetation 
which does not transpire large quantities of water, must inevit- 
ably affect the climate, not so much the climate of the region 
in which the destruction took place but the drier regions into 
which the prevailing air-currents flow." 
Finally, after having weighed what many writers have written, 
the preponderance of evidence seems to show that the forest 
acts as a " governor " of stream flow, rather than as a means 
of increasing precipitation. 
Erosion following forest destruction may be due to either 
wind or water, and may be great or small, depending upon 
climate, depth and character of the exposed soil and its slope. 
Col. H. M. Chittenden in a very full paper f claims that " soil 
erosion does not result from forest cutting in itself, but from 
cultivation, using that term in a broad sense." 
Such a conclusion will certainly not hold true in all districts, 
for instance in the Canadian locality shown on page 301. The 
photograph was kindly sent the author by Mr. Elwood Wilson 
of Grand Mere, P. Q. 
Having once carefully selected a watershed, it should be 
protected with the greatest care which science suggests, and 
* Science, July 18, 1913. 
J Proc. Am. Soc. C.E., Sept., 1908. RIVER AND STREAM WATER 
301 
SHOWING EROSION FOURTEEN YEARS AFTER REMOVAL OF THE Ttmtiep 302 
WATER-SUPPLY 
with the utmost vigor which the law allows. Right here is the 
weakness shown by many of our city councils. The law is 
strong enough, and the municipal rights are plenty, but it is 
often very difficult to move the authorities to proper action. 
A most fruitful source of evil arises from the unquestioned 
right of a riparian landholder to " water his stock." The 
broad interpretation of this right can be carried to an absurd 
degree; for instance, the writer has seen the open channel-way 
connecting the storage and distributing reservoirs of a large 
city doing duty, at one point of its course, as a farmyard 
drain, the cattle standing in the small stream at pleasure. 
It may not be amiss to point out that regulations for the 
protection of a watershed which do well enough during sum- 
mer months may entirely fail of effectiveness after the ground 
becomes frozen. Drainage material which at one time could 
sink into the ground and become oxidized by infiltration 
would at other seasons flow down steep slopes over the frozen 
surface, or, if itself arrested by frost, would be at a latter date 
washed into the stream by melting snows over the yet un- 
thawed ground. Such cases of contamination are not rare? 
and may be followed by most serious consequences, as was 
instanced by the outbreak of typhoid fever at Plymouth, Pa. 
It is useless to depend upon the purifying action of frost, for, 
as has been shown, typhoid germs can withstand being frozen 
in solid ice during a considerable period. 
As further reasons why winter conditions may increase the 
danger due to a polluted water, let it be noted that: Adverse 
organisms are fewer; sunlight is weaker; viscosity of the 
water being greater the sedimentation is slower; and the tem- 
perature of the water being lower conduces to greater typhoid 
longevity. 
To repeat what we have already touched upon, the purify- 
ing action of filtration through common soil is a point fre- 
quently misunderstood, and yet of important bearing when 
considering the protection of a watershed. Such filtration is 
only effective when it is intermittent. 
The nitrifying organisms which accomplish the oxidation RIVER AND STREAM WATER 
303 
of the objectionable sewage material can only operate in pres- 
ence of atmospheric oxygen. A supply of air must be present 
in the pores of the soil or else oxidation ceases. After a " dose " 
of sewage has been applied to a soil a sufficient interval must 
elapse to permit the air to renew the exhausted oxygen; other- 
wise the slow-moving and continuous stream of filth must 
carry its objectionable properties to considerable distances. 
How important, then, that every privy located within 
drainage distance of a source of water supply should be built 
without a vault, and should have its cleanings removed at fre- 
quent intervals and disposed of in a safe manner. 
Finally, however desirable for the moment a river may be as 
a source of water supply, it must not be forgotten that the con- 
ditions may change in the course of years with the growth of 
population up-stream, as has been already noted on another 
page. Objection is properly raised to the use of the unfiltered 
water of a large river, on the ground that pollution of the stream 
by sewage material is certainly on the increase, and that the 
introduction of sewage systems in the towns above will, at no 
distant date, render the river-water very undesirable. 
It has to be admitted that the water of even a mountain 
stream, if taken in the raw state, might readily prove itself 
more objectionable than that of a large though polluted river, 
if the latter were efficiently filtered before delivery to the 
consumers, for the reason that three or four families on the 
small stream might, unknown to the city authorities, chance to 
have their privies and drains empty directly into the brook 
without soil intervention. 
Further consideration of care of water sheds will be found 
under " stored water." CHAPTER VII 
STORED WATER 
Nature provides enormous quantities of water stored up 
in lakes and ponds ready for human consumption, and man 
frequently supplements this provision by impounding waters 
in artificial basins when the natural reservoirs of the district 
are unavailable or are insufficient in size. Some of these 
artificial lakes are of great extent, thus the Ashokan reservoir 
for New York City has an area of about thirty-six square 
miles with an average depth of fifty feet. Still larger are 
the St. Maurice Reservoir (Quebec), capacity 160 billion 
cubic feet; Ga tun Lake (Panama), capacity 183 billion cubic 
feet; and the reservoir at Assuan (Egypt), 3750 billion cubic 
feet. 
Lakes of such great size as to be properly considered inland 
seas-the Great Lakes of North America, for instance-furnish 
water of quite constant composition, free from the considerable 
vegetable contamination so frequently met with in small lakes 
and ponds. 
Lake Superior, 
near Duluth, 40 
Miles Out from 
Shore. 
March, 1896. 
Lake Michigan* 
12 miles out 
from Chicago. 
Oct. 23, 1896. 
Lake Erie, near 
Erie, Pa., 14 
miles out 
from Shore. 
October, 1897. 
Free ammonia 
.03 
0.0 
.045 
Albuminoid ammonia 
.02 
.08 
.112 
Chlorine 
2. 
5 • 5 
3 ■ 5 
Nitrogen as nitrates 
.00 
0.0 
.08 
Nitrogen as nitrites 
0.0 
0.0 
trace 
Required oxygen 
i-i5 
1.6 
1 25 
Total solids 
54- 
130. 
134- 
* " Streams Examination," p. 18. 
304 STORED WATER 
305 
MINERAL ANALYSES OF WATERS OF THE GREAT 
LAKES * 
(Average of io or more analyses for each (by Dole) 
Parts per Million. 
Superior. 
Huron. 
Michigan. 
Erie. 
River St. 
Lawrence. 
Samples taken at 
Sault 
Ste. 
Marie. 
Port 
Huron. 
St. 
Ignace. 
Buffalo. 
Ogdens- 
burg. 
Turbidity 
Silica (SiO2) 
Iron (Fe) 
Calcium (Ca) 
Magnesium (Mg) 
Sodium and potassium (Na+K). 
Carbonates (CO3) 
Bi-carbonates (HCO3) 
Sulphates (SO4) 
Nitrates (NO3) 
Chlorides (Cl) 
Total solids 
Total hardness (calculated) 
2. 
7-4 
.06 
13- 
3-i 
3-2 
0.0 
56. 
2.1 
•5 
1.1 
60. 
45-4 
tr. 
12 . 
.04 
24. 
7- 
4-4 
1.8 
100. 
6.2 
•4 
2.6 
108. 
89.2 
tr. 
10. 
•04 
26. 
8.2 
4-7 
2.9 
112. 
7-2 
•3 
2-7 
118. 
99.2 
41 • 
5-9 
•07 
31- 
7-6 
6-5 
3-i 
114. 
13- 
•3 
8-7 
*33 • 
108.2 
4-5 
6.6 
05 
31- 
7-2 
6-3 
2.9 
116. 
12. 
•3 
7-7 
134- 
107.5 
Large as these Great Lakes are, the influence of the sewage 
from cities upon their shores is nevertheless beginning to be 
seriously felt. The pollution of Lake Michigan by the sewage 
of Chicago is a widely known fact, and the intakes, situated 
as they are one to five miles from shore, are frequently reached 
by the ever swelling volume of the city's refuse. 
It was with the object of preventing the pollution of 
Chicago's water supply that the Chicago Drainage Canal was 
constructed, whereby the current of the Chicago River was 
reversed and the city's sewage was washed into the Mississippi 
valley by water from Lake Michigan. 
" The canal was put in operation in January, 1900. It 
was designed to have a capacity of 10,000 cubic feet per second, 
with a velocity of 1.25 miles per hour in earth and 1.9 mile per- 
hour in rock. The capacity was based on the assumption 
that in order to prevent nuisance it was necessary to dilute 
the sewage to the extent of providing 3I cubic feet per second 
for each 1000 persons discharging sewage into the canal. 
* J. N. E. Water Works Assn., 23: 259 306 
WATER-SUPPLY 
" The canal has a total length of 28.05 miles and a minimum 
water depth at low stage of Lake Michigan of 22 feet. The 
rock section is 15.95 miles long, 160 feet wide at the bottom 
and 162 feet wide at the top. A portion of the earth section, 
5.3 miles long, is 202 feet wide at the bottom and has side 
slopes of 2 to 1. The remainder of the canal, also in earth, 
is 7.8 miles long, and as originally built was no feet wide at 
the bottom, with side slopes of 2 to 1." 
Effort has been made to have the amount of water ab- 
stracted from Lake Michigan increased, but the report of the 
International Waterways Commission was adverse to this 
movement, their recommendation being " that the Govern- 
ment of the United States prohibits the diversion of more than 
10,000 cubic feet per second for the Chicago Drainage Canal." 
Much opportunity is given in large lakes for sedimentation 
to come into full play, and settlement is, in consequence, a 
great item in the process of the natural purification of their 
waters. 
So far as the American " Great Lakes " are concerned, 
there is a difference in this respect between Lake Erie and 
the other members of the group. Lake Erie is comparatively 
shallow and is stirred to its bottom by every gale, therefore 
sedimentation advantages therein are at a minimum. 
Naturally the shallow water near the shore of a lake would 
show the greatest amount of bacterial life. It would also be 
expected that the advantage of sedimentation would be made 
more evident in the character of the water of the outlet than 
in that of an entering stream. 
Dunant found 150,000 bacteria per cubic centimetre in water 
from Lake Geneva taken near the shore, and only 38 per cubic 
centimetre in a sample from the middle of the lake. Percy 
Frankland examined the waters of two inlets of Loch Lin- 
trathen and found them to possess counts of 1700 and 780 
per cubic centimetre respectively, while the outlet of the loch 
contained but 30 per cubic centimetre. 
In smaller lakes and ponds the influence of vegetation begins STORED WATER 
307 
to be felt, sometimes so seriously as to interfere with the use 
of their waters for potable purposes, and we are confronted by 
the question, what, if any, are the advantages and disadvantages 
of reservoir storage. 
Tersely stated, the advantages of water storage are many 
and the disadvantages but few. In those days when the ex- 
pression " stagnant water " carried with it all sorts of ill- 
defined fears, the opponents of storage were easy to find, and 
their enthusiastic statement that " abundance of light and 
air is essential to the proper conditioning of water for human 
consumption " received very general support. He who drank 
of the rapid stream was accounted greater in wisdom than he 
who selected a less aerated supply. 
It is now almost trite to say that still water, rather than run- 
ning water, purifies itself the better, and it equally lacks novelty 
to point out that the more rapidly a stream flows, the sooner is 
its load of pollution delivered to the thirsty consumer. In that 
connection, let it be again said that a great deal more depends 
upon the number of hours required for stream-flow than upon 
the distance in miles between the intake and the source of 
pollution, and a concise statement giving information as to 
time of flow should appear in a report covering the sanitary 
survey. 
Stagnation has its disadvantages, of course. Increase in 
color naturally follows if water be permitted to long remain in 
contact with a muddy bottom loaded with soluble extractive 
matters. Not only is damage to the water's physical appear- 
ance a result of such contact, but the material passing into 
solution is likely to furnish abundant food for those minute 
forms of life which carry objectionable tastes and smells to many 
public waters. 
To the layman's ear the word " stagnant " has a most 
unpleasant sound, intimately associated with the production 
of disease; and yet its origin is innocent enough, viz., " stag- 
num," a piece of standing water, not running in a current or 
stream. 
Pools so overloaded with vegetable growth, both dead and 308 
WATER-SUPPLY 
alive, as to be unfit for human drinking are plenty and they 
are commonly stagnant, but the ill-favored word does not prop- 
erly apply alone to those; it is just as applicable to a water of 
crystal clearness resting upon a bed of sand. 
The condition of " standing water " just referred to,- 
namely, that of being overstocked with vegetable growths,-is 
practically the only one toward which objection can point when 
considering the pros and cons of reservoir storage. 
The word " vegetable " is here to be taken in its broad sense, 
as it should include not only those growths which would be 
recognized by the public at large, but also those of the minute 
world as well, which latter constitute one branch of that lake 
life known as " plankton." 
As one of many instances of excessive overgrowth, a small 
lake could be named which is so loaded with dense vegetation 
that decay gets ahead of new growth, and the use of its water 
is productive of temporary diarrhoea. It is a stained water, 
but its color is not to be considered as a measure of its objec- 
tionable qualities for table use, as many waters of much darker 
tint are of excellent quality for such purpose. Color and 
fitness for drinking bear no relation to each other. 
It is true that colorless waters are now demanded by the 
people, and " meadow teas " are growing in disfavor, but that 
change in public opinion is of recent date and is not based upon 
sanitary considerations. 
The deepening in color of the lower layers of a water stored 
upon an unclean bottom; the encouragement of growth of small 
organisms producing taste and smell by reason of an accumula- 
tion of extracted food suitable for their development; and a 
deficiency in dissolved oxygen in the bottom levels, constitute the 
sum of objections that can be raised to the impounding of water, 
and they are much more than balanced by the advantages 
that accrue from such storage. 
Observations have been made by W. A. Manss of this 
laboratory, showing the influence of depth upon the character 
of water in a storage reservoir of about 3000 areas. 
He found that during the summer season, owing to the STORED WATER 
309 
comparative lightness of the warmer water, no circulation 
takes place below a depth of twenty-five feet, that being the 
usual distance to which wind and wave agitation extends. 
Should a lake be protected from the wind, the aerated layer 
may extend temporarily from ten to forty feet from the sur- 
face, depending upon the extent of the water-surface, and below 
this level the cold, stagnant water rests, until such time as the 
chilling of the upper layer increases its gravity to and beyond 
that of the lower layer upon which it floats. When this point 
is reached, readjustment of relative position is immediately 
instituted, in accordance with the change in specific gravity, 
and the water of the lake " turns over." 
The formation of this stagnant layer begins in April in this 
latitude, and circulation is partly re-established in October and 
completely so in November. With the advent of freezing 
weather a second period of stratification is inaugurated which 
continues until the surface thaws again in the spring. Vertical 
circulation then progresses until the warm sun of later April 
renders the surface-water so light as to float upon the colder 
layers beneath, when summer stagnation again begins. 
Whenever the lower stagnant layer is brought in contact 
with the decomposing organic matter, as is the case in reservoirs 
with bottoms from which the vegetation has not been removed, 
the dissolved oxygen present is quickly used up; quantities of 
extractive matters pass into solution and the water acquires 
an odor and becomes dark in color. 
Many analyses are available showing this diminution and 
ultimate total exhaustion of dissolved oxygen in the stagnant 
layer. 
Even though the bottom of a lake or reservoir be perfectly 
clean and sandy, the dissolved oxygen must surely diminish 
to some extent in the lower layers of the water. 
That the stagnation of water in the lower levels of a reser- 
voir is not in itself obj'ectionable has been shown by Dr. 
Drown in his study of reservoirs Nos. 3 and 4 of the Boston 
Water-works: 
11 Water in the stagnant layer does not become foul unless 310 
WATER-SUPPLY 
there is decomposable organic matter present. Thus in Basin 
4 of the Boston Water-works, which was carefully prepared 
for the reception of the water by the removal of all soil and 
vegetable matter, and is supplied with a brown, swampy water 
from a watershed almost entirely free from population, the 
water is good at a depth of forty feet, because the water con- 
tains very little organic matter with a tendency to decomposi- 
tion. 
RESERVOIR NO. 3, BOSTON WATER-WORKS (AUG. 20, 1891) 
Temperature, Per Cent of Dis- 
Deg. Fahr, solved Oxygen. 
Surface.. . 
74-7 85.88 
6 feet below surface 74.7 85.06 
12 " 
14 
70-9 58-97 
" 0 
15 " 
17 " 
19 
11 0 
" 0 
" 0 
21 
(bottom)... 62.8 0 
RESERVOIR NO. 4, BOSTON WATER-WORKS (AUG. 20, 1891) 
Surface.. . 
74-7 
84-5° 
io feet below surface 
• 70-9 
84.42 
20 
c c 
. 61.9 
28.02 
30 
( c 
27.42 
35 
C ( 
• 54-7 
16.28 
361 " 
(bottom). 
• 54-7 
15.10 
" The contrast in the condition of the water in these two 
reservoirs is very striking. Reservoir No. 3, in which the 
oxygen is exhausted at a depth of 14 feet, receives a not in- 
considerable amount of direct pollution from the towns of 
Marlborough and Southborough, while the drainage-area of 
Reservoir No. 4, as has been already said, is very sparsely 
poplulated." 
Uniform experience goes to prove that good water may be 
preserved in properly constructed reservoirs without deteriora- STORED WATER 
311 
tion for indefinite lengths of time. It must be remembered, 
however, in this connection, that to keep a ground-water in 
good condition it is necessary to cover the reservoir. Such 
waters are usually charged with mineral matter suitable for 
plant-food, and the growths of algae will be very likely to 
appear therein unless light be excluded. Thus the great reser- 
voirs supplying the spring-waters of Paris are kept entirely 
dark, with the best of results. Reservoirs used to store fil- 
tered surface-waters should be likewise covered for the same 
reason. 
Algae, and directly or indirectly the " plankton " in general, 
largely depend for their development upon material furnish- 
ing nitrogen. Water containing a moderate amount of this 
element washed from natural sources will sustain but a small 
growth of such life, but where nitrogen is present in great 
quantity as nitrates, as is the case in many deep-seated waters, 
the development of algae is often excessive during storage, 
if light be admitted. 
The effluent from a well-designed sewage-disposal plant 
would, if added to stored water, cause a heavier growth of algae 
than would be produced by the raw sewage itself because of its 
containing a larger quantity of nitrates. 
It must not be sweepingly assumed that all the " Plankton " 
life is to be rated as uniformly objectionable; quite the con- 
trary, as a reasonable degree of it acts as a distinct help in main- 
taining the safety of natural waters. Thus we find " bacteria- 
eaters " such as many kinds of ciliated infusoria, rotifers, 
daphnia and the like, feeding upon minute germ life, and doing 
so to our great advantage. 
To quote from a translation by Kuichling: 
" The question is, what becomes of the great quantities of 
offal and excreta, the many remnants of decaying plants, the 
refuse of communities, and the finely divided factory wastes 
of every description, which find their way into our streams, 
even under normal conditions, if a large portion thereof is not 
consumed by the aquatic detritus-eaters and the omnivorous 
fauna before settling to the bottom." 312 
WATER-SUPPLY 
An investigation by C. P. Hoover * shows that " Intestinal 
organisms will not live in water containing no free or half- 
bound carbonic acid. 
Lime-softened water is fatal to bacteria of the colon and 
typhoid group in forty-eight hours. The action is selective 
in that certain harmless bacteria grow but the disease-producing 
germs do not." 
As algal growths use up dissolved carbon dioxide it is evi- 
dent that this form of plant life will act as a protection against 
such pathogenic bacteria as may find admission to storage 
reservoirs. 
Experience indicates that storage of surface-waters in open 
reservoirs causes but little change in the character of such 
waters, and that what small change does take place is bene- 
ficial, but the bad effect of the open storage of ground-waters 
is quite strongly marked. No better illustration could be given 
of the different actions of surface and ground waters during 
similar conditions of storage in open reservoirs than to call 
attention to the results secured at Atlantic City, N. J. The 
two supplies of the city, although afterwards mixed, are stored 
separately in basins of about the same size, separated one from 
the other by an embankment only a few feet wide. See illus- 
tration on page 313. The water shown in the foreground 
is from flowing wells which deliver through the bottom of the 
reservoir. It is continually filled with large masses of green 
growth. The reservoir seen in the background holds surface- 
water, admitted by canal, and it is entirely free from vegetation 
of any kind. A more instructive instance of varied results 
under identical storage conditions could hardly be asked. 
To put it tersely, it may be said that a water from a dark 
source should be stored in the dark and vice versa- filtered 
water being the exception, as above noted, and requiring dark 
storage irrespective of its source. 
The chart on page 314 which graphically shows the tem- 
perature variations during the summer season for different 
depths of Lake Cochituate, was prepared by Desmond Fitz- 
* Engr. Record, Sept. 6, 1913. STORED WATER 
313 
Gerald for the report of the Boston Water-works. It will be 
noted that the temperature curves run together at the times of 
the semi-annual 11 turn-over." 
As supplementary to his investigation concerning the 
amount of dissolved oxygen in the water of ponds and reser- 
voirs at different depths during the summer months, Dr. Drown 
made similar determinations during severe winter weather. 
VIEW OF WELL-BASIN, IN THE FOREGROUND, AND CANAL BASIN, IN THE BACK- 
GROUND, AT ABSECON PUMPING STATION, ATLANTIC CITY, N. J. 
when the waters in question were covered with thick coatings 
of ice. The winter results fully confirmed those of summer, 
and showed that with exclusion of air the dissolved oxygen 
diminished in proportion to the quantity of organic material 
present.* 
As a result of the 11 turning over " in the spring and autumn 
the waters of iakes and deep reservoirs become fouled to a 
* Rep. Mass. Board of Health, 1892, p. 331. 314 
WATER-SUPPLY 
TEMPERATURE S STORED WATER 
315 
greater or less degree throughout their entire masses by virtue 
of the mingling of the waters of all layers during these periods 
of vertical circulation. 
The deeply stained water of the bottom imparts a shade 
of its color to the body of the water at large, and the nitro- 
genous matter in solution, oxidizing to " nitrates," furnishes 
food for countless millions of " diatoms," and other growths 
whose development and decay cause many of the unpleasant 
tastes and odors with which our city supplies are so frequently 
afflicted. 
Tastes and odors occur in waters of streams and rivers as 
well as in those of lakes, but the instances are fewer and the 
intensity is not so great. 
The Boston Water-supply Department has made extended 
study of the coloring-matter common to the stagnant layer, 
and of the observed facts that the color at first deepens on 
exposure to air and afterwards bleaches out. The department 
finds that these phenomena are more strongly marked in pro- 
portion as the bottom-water is rich in salts of iron and man- 
ganese. 
Those familiar with the properties and behavior of ferrous 
and ferric salts would have predicted that the soluble and light- 
colored ferrous compounds would, upon exposure to the atmos- 
pheric oxygen, oxidize to darker ferric salts, and ultimately 
fall as insoluble hydrated oxide and carbonate, leaving the 
water bleached; and it may be added that the slow fall of this 
depositing iron aids in ridding the water of other foreign 
materials, very much in the manner that the 11 coagulants " 
assist the purification process in mechanical filtration. 
Sundry vegetable and peaty extracts are exceedingly diffi- 
cult to decolorize, and waters containing them cannot be ren- 
dered colorless by storage in presence of light and air in 
a period short of many months. An improvement in color 
always results from open storage, but its entire removal is often 
impossible. 
FitzGerald reports the following seasonal changes in color 316 
WATER-SUPPLY 
of the waters of the Sudbury: The highest color is attained in 
the month of June, and then it rapidly lessens until September. 
Towards the end of October the color increases again until 
December, and then decreases until it reaches its yearly mini- 
mum in the middle of March. He offers the following 
explanation: In the early spring the swamps are over- 
flowed and the color is low on account of dilution. Concen- 
tration causes increase in color until early summer, after 
which time the swamp pools cease to overflow, and conse- 
quently the brooks grow clearer. Autumn rains again fill the 
deeply colored swamp pools to overflowing and wash highly 
stained water into the streams, increasing their color, which 
is afterwards lessened in winter by the freezing up of the swamp 
sources.* 
Subsequently FitzGerald experimented regarding the action 
of lights of different colors on the reduction of the brown color 
of water. The water under experiment was exposed during 
one month of summer in bottles of colored glass. 
Original 
Color. 
Final 
Color. 
Per Cent 
Reduction 
in Color. 
In white bottle 
I-O5 
°-39 
c c c c 
O.85 
0.19 
Mean 
°-95 
0.29 
69.48 
In blue bottle 
I.O2 
°-39 
• 
Cl Cl 
O.85 
0.24 
Mean 
°-93 
0.31 
66.66 
In yellow bottle...... 
1.02 
o-58 
IC CC 
0-85 
0.46 
Mean 
°-93 
0.52 
44-09 
In red bottle 
1.02 
o-54 
Cl c c 
0-85 
o-47 
- 
- 
Mean 
0.93 
o-5° 
46.24 
* " Met o. Water-supply," Rep. Mass. Board of Health, 1895, Appendix 3. STORED WATER 
317 
DIAGRAM SHOWING THE RELATIONS BETWEEN THE DIATOM 
GROWTHS AND THE STAGNATION AND CIRCULATION OF THE 
WATER IN LAKE COCHITUATE. 
I DIATOMS. 
Average number per c.c. 
for the surface, mid-depth, 
and bottom. 
COLOR. 
(Nessler Scale) 
(After Whipple) 318 
WATER-SUPPLY 
To return to what has been said concerning the growth 
of diatoms, the chart on the page opposite, prepared by Prof- 
G. C. Whipple,* makes the relation of such growth to the periods 
of vertical circulation very distinct. 
Among other conclusions he found that diatoms flourish 
best in ponds having muddy bottoms; that their growth is 
directly connected with the phenomenon of stagnation; that 
their development does not occur when the lower strata of 
water are quiescent, but rather during those periods when the 
water is in circulation from top to bottom; that the' two 
most important conditions for their growth are a sufficient 
supply of nitrates and a free circulation of air; and that both 
these conditions are found during the periods of vertical cir- 
culation. 
Light also stimulates their growth. Hence it was to have 
been expected that a darkly stained water, which does not 
easily admit light, would prove a less favorable medium 
for the propagation of diatoms than a water of colorless 
character. 
As has been said, air is essential to the abundant growth 
of these minute organisms, but the introduction of such air by 
means productive of violent agitation of the water would seem 
to be quite fatal to their development; probably through direct 
damage to the fragile cells of the plants. 
Whipple observed the bad effect upon the growth produced 
by blowing a current of air through the water, f As a deduc- 
tion from this we note a benefit to be derived from reservoir 
fountains and other means of agitation. 
It has also been shown J that many organisms of fragile 
structure are broken up by the pressure and currents of 
the mains, and that others die therein from lack of light 
and food. The oil they contain (see page 14) is thus liber- 
ated by disintegration, and as a result the water of the 
* " Observations on the Growth of Diatoms in Surface-waters." 
f J. N. E. Water-works Asso., xi. 3. 
I Ibid., xii. 7. STORED WATER 
319 
mains, as drawn at the taps, may smell worse than that of the 
reservoir. 
It must be noted that the same organism may produce 
two very different odors, one during its period of development 
and quite another one at the time of its decay. For in- 
stance, an abundant growth of anabana causes a distinct smell 
of " green corn " while the plants are yet alive, which changes 
to a pronounced 11 pig-pen " odor during the process of their 
decomposition. This change is quite characteristic. 
While speaking of anabcena let it be added that the author 
observed a very interesting growth of this organism in the 
reservoir of a Western city. 
The plant was distributed throughout the upper level from 
the surface to a depth of ten feet, producing disagreeable taste 
and smell in that part of the water, but leaving the deeper 
portions unaffected and entirely fit for use. Because of a 
sudden change of temperature, the water of the reservoir 
" turned over " in a single night, thereby distributing the 
organism and admitting highly objectionable water to the 
mains. Widespread public complaint was, of course, the im- 
mediate result. 
This instance but adds weight to the demand for such form 
of gate-house in all city reservoirs as will permit of tapping off 
the water at any desired level; and it also shows, not only the 
necessity for biological and temperature observations, but that 
the entire storage need not be cut off from use because of the 
lack of fitness of a portion of the supply. 
One of the most instructive cases of reservoir contamination 
by aquatic growth occurred at Brooklyn, N. Y., in 1896, and 
is reported by Leeds and Whipple in Engineering News, July 1, 
1897. The cause of the difficulty was an abundant develop- 
ment of asterionella. 
Judging from analysis of the bodies of asterionella, Whipple 
and Jackson found that in order to support a growth of 
such diatoms the water should contain the following, in parts 
per million: 320 
WATER-SUPPLY 
N as NO3 079 
SiCL 1.780 
Fe2O3 083 
CaO 052 
MgO 045 
K2O 043 
MnoOs 030 
SO3 014 
Total solids.. . 3.600 
They considered the silica, iron, and manganese as those items 
most likely to be lacking in quantity sufficient to support an 
abundant growth of the plant.* 
As is well known, the city of Brooklyn uses a mixed water, 
part of it being derived from wells driven into the sandy soil 
of Long Island and the balance of it coming from surface 
sources, such as Jamaica pond. The surface-water supplied the 
asterionella seed, the ground-water the necessary food, and the 
open Ridgewood reservoir, where the mixed waters were stored, 
allowed the light and heat of the summer sun to encourage 
abundant growth. The remedy in this case lay in using a by- 
pass direct to the mains, until such time as the reservoir water 
had recovered its normal quality. Light is essential to such 
a growth and the use of a by-pass was cheaper than constructing 
a reservoir cover. 
A similar outbreak occurred at Penzance, England, in 
1904. As at Brooklyn it arose from mixing a surface water 
with that from a deep well. J 
ORGANISMS AND THE NUMBER OF THEM PER C.C. REQUIRED 
TO PRODUCE ODOR OR TASTE 
Reported by Whipple J 
Number required per c.c. 
Character of Odor 
and Taste. 
Synedra pulchella 
5000 
earthy-vegetable 
Cyclotella 
5OOO 
aromatic-fishy 
Melosira 
over 3000 
earthy-vegetable 
Anabama. 
1700 standard units 
moldy grass, green corn 
Scenedesmus. 
25000 standard units; 
vegetable-aromati c 
* J. N. E. Water-works Asso., xiv. 19. 
t La Technique Sanitaire, June, 1906. 
t See also Engineering News, June, 7 1900. STORED WATER 
321 
As to asterionella Whipple finds, by experiment, the fol- 
lowing relations between the aromatic, fishy odor of the water 
and numbers of the organism present per cubic centimetre:* 
Odor. Number per c.c. 
None o to 1,000 
Very faint 500 to 3,000 
Faint 1,000 to 5,000 
Distinct ' 3,000 to 15,000 • 
Decided 10,000 and over 
In a letter to the author Professor Whipple gave the table 
on page 322 showing numbers of organisms found in cases of 
observed odor. 
He added: " As to stating the number necessary to cause a 
noticeable odor, I do not feel sure enough about the figures to 
quote any. It might be done approximately in a few cases, 
as with asterionella, but I am sure that we cannot do more 
than guess at most of them." 
As to what could be done to rid a water of the odors due to 
algal and other growths, the outlook was not very hopeful 
previous to the appearance of the " copper sulphate process " 
proposed by Moore and Kellernan in 19'04 (Bulletins 64 and 76, 
U. S. Bureau of Plant Industry). 
The method of applying this chemical is simple enough, and 
its use is very efficacious. Bought in bulk, it can be had at 
about five cents per pound, and its distribution is readily se- 
cured by filling it into perforated buckets, or even bags, and 
towing the same by row-boat or launch over the reservoir 
surface. 
Decided objection was raised against such a process of 
" disinfecting " a public water supply, and the opposition was 
expecially marked in England, but the use of it is still with us 
and is likely to stay, for the reason that the " dose " is minute 
* J. N. E. Water-works Asso., xiv. i. 322 
WATER-SUPPLY 
Organism. 
Odor. 
Date. 
Locality. 
Number of 
St. Units 
per c.c. 
Asterionella 
Aromatic, geranium, fishy 
Dec., 1897 
Mt. Prospect Res., Brooklyn 
50,000 
Cyclotella. . 
Faint aromatic 
April, 1899 
Ridgewood Res., Brooklyn 
25,000 
Diatoma 
c c cc 
Oct., 1900 
Mt. Prospect Res., Brooklyn 
13,000 
Meridion 
Aromatic 
May, 1894 
Pond at Chestnut Hill, Mass. 
5,000 
Tabellaria 
c c 
May, 1890 
Lake Cochituate 
2,500 
Melosira 
Vegetable 
Oct., 1900 
Hempstead, L. I. 
8,000 
Synedra 
C c 
Sept., 1897 
Ridgewood Res., Brooklyn 
20,000 
Vol vox 
Fishy 
Faintly fishy 
c c 
Rochester, N. Y. 
Mt. Prospect Res., Brooklyn 
Eudorina 
Aug., 1900 
900 
Fandorina 
Aug., 1895 
Pond at Chestnut Hill, Mass. 
5,000 
Dictyosphaerium 
C c 
Aug., 1897 
Mt. Prospect Res., Brooklyn 
1,000 
Anabaena 
Rivularia 
Grassy and moldy, green com, etc. 
Grassy and moldy 
Aug., 1899 
Long Pond, N. J. 
9,000 
3,ooo 
Aphanizomenon 
Grassy 
Oct., 1896 
Laurel Lake, Fitzwilliam, H. N. 
3,ooo 
Uroglena. ... 
Fishy and oily 
April, 1892 
Norwood, Mass. 
17,000 
Synura 
Cucumber 
Oct., 1891 
Glen Lewis Pond, Lynn, Mass. 
7,000 
Dinobr^on 
Bursaria 
Fishy and rockweed 
Fishy, Irish moss 
July, 1890 
Scott Res., Fitchburg, Mass. 
4,5oo 
Peridinium. . . 
Fishy, clam-shells 
July, 1891 
Mystic Lake, Somerville 
7,5oo 
Glenodinium 
Fishy 
Aug., 1893 
CC C C 
10,000 
Chlamydomonas 
Fishy and moldy 
Nov., 1898 
Spot Pond, Malden, Mass. 
700 
Cryptomonas 
Aromatic 
Jan., 1900 
Creek, Lodi, N. J. 
10,000 
Mallomonas 
Aromatic, fishy 
Sept., 1896 
Lake Cochituate 
3,5oo 
NUMBERS OF ORGANISMS FOUND BY WHIPPLE IN INSTANCES OF OBSERVED ODOR STORED WATER 
323 
and is only occasionally required, that it is reliable in results, 
and that experiment has shown that it is not followed by the 
evil consequences predicted. 
It must be remembered that it is not added to the water 
continually, but is used only at stated and widely separated 
intervals,-namely, at those times when the " crop " of minute 
organisms become so well grown as to produce objectionable 
effect upon the water. 
Even though all of the chemical employed reached the 
public water-mains, which is far from being the case, yet the 
actual quantity of metallic copper so represented would bear 
favorable comparison with the amount of the element present 
in many of our common articles of food; for instance: 
QUANTITY OF COPPER, CALCULATED AS THE METAL, IN VARIOUS 
FOODS. STATED AS MILLIGRAMS PER KILOGRAM * 
Almonds 36.8 
Apricots 1. o 
Cherries 2.3 
Cocoa 47.0 
Cucumbers 45. o 
Egg, yolk 5.6 
Egg, white 7.2 
Figs 15.1 
Grapes 1. o 
Kidney (beef) 4.0 
Milk (cow) 1.6 
Oatmeal 4.2 
Peas (French) 59.4 
Potatoes < 2.8 
Strawberries 8.0 
The Tomhannock reservoir at Troy, N. Y., is occasionally 
li coppered " with a dose of 1 part of bluevitriol (copper sulphate) 
to 3,500,000 parts of water, by weight. This would corre- 
* J. Ind. and Engr., Chem., 7: 498. 324 
WATER-SUPPLY 
spond to .073 milligram of metallic copper per kilogram of 
water. 
According to Kemna it must be admitted that the salts 
of copper are poisonous in certain doses, but it is equally true 
that their toxicity has been greatly exaggerated, and the pro- 
scription of traces of the metal in foods has been one of the 
errors of hygiene. 
The human organism easily tolerates copper and the 
small quantities employed to destroy algas are absolutely 
harmless.* 
Perhaps one reason why the " coppering." of reservoirs has 
led to so much criticism is because of the dead fish that are to 
be seen after the chemical has been applied. When considering 
this effect upon fish life, one should bear in mind that the 
11 dose " has of necessity to be applied uniformly over the sur- 
face of the water, and each acre of such surface presumably 
receives the same amount, irrespective of the depth of water that 
the acre covers. As a result, the shallow parts of the lake 
receive temporarily a greater quantity of the sulphate, per cubic 
foot of water, than do those which are deeper; again, the entire 
quantity of chemical intended for the whole body of the lake 
is delivered to a few inches of its surface layer; therefore, 
until diffusion has taken place, fish which chance to swim into 
such water receive a very concentrated dose and are likely to 
be affected by it. Distribution is complete by the time the water 
reaches the public 'mains, and, moreover, the minute dose 
used has been more or less completely disposed of through its 
action upon} the organisms for whose destruction it has been 
employed. 
By a judicious watching of the shallow bays of a reservoir, 
from which seeding of the whole main body of water may at 
times occur, and by applying the sulphate of copper to these 
bays locally, the need for the chemical throughout the reservoir 
may often be avoided. For such work the use of a syringe 
such as is employed in a green-house for the spraying of plants 
is very convenient. 
* La Tech. San., Aug. 1906, p. 172. STORED WATER 
325 
In a paper before the Section on Hygiene of the Eighth 
International Congress of Applied Chemistry, 1912, Kellernan 
presented the following table: 
QUANTITY OF COPPER SULPHATE REQUIRED TO KILL VARIOUS 
FORMS OF ODOR-PRODUCING ORGANISMS 
Copper Suplhate Required, Expressed as Parts, per Million Parts of 
Water 
Anabaena 09 
Asterionella 1 
Beggiatoa 5. 
Chara 2 to 5. 
Cladophora 1. 
Cladothrix 2 
Clathrocystis 1 
Ccelosphaerium 3 
Conferva 4 to 2. 
Euglena 1. 
Fragilaria 25 
Hydrodictyon 1 
Kirchnerielh 5. to 10. 
Leptomitus 4 
Microspora 4 
Navicula 07 
Oscillatoria 1 to .4 
Peridinium 2. 
Scenesdesmus 5. to 10. 
Spirogyra 05 to .3 
Ulothrix 2 
Uroglena 05 
Volvox 25 
Zygnema 7 
He adds a list of twelve genera of algae that in his experience 
are causing trouble in reservoirs and ponds: 
NUMBER OF OBSERVED CASES 
Anabaena 27 
Asterionella 9 
Beggiatoa 20 
Chara 26 
Cladophora 17 
Clathrocystis 23 
Conferva 56 
Crenothrix 13 
Fragilaria 19 
Navicula 21 
Oscillatoria 49 
Spirogyra 43 
Jackson claims that blue-green algae will die if the water be 
" coppered " one part to five million. His dose for Mellosira 
or Synedra is one to two million, and he claims that the former 
gives no odor of growth, but only that of decay. He finds 
that coppering runs out certain forms of organisms and substi- 
tutes others by a sort of selective action, but those thus sub- 
stituted are not likely to be odor-producers; and he further 
notes that, while " bottom " or decomposition odors are easily 
shaken out by aeration, " top odors, viz., those of growth, 
have to be removed by filtering out the organism, or killing 
them by copper sulphate, or both. In this opinion, filtra- 326 
WATER-SUPPLY 
tion of either type is effective for removal of odors of growth, 
but he believes that aeration would be worse than useless for 
living plankton, for the reason that the agitation would tend 
to mechanically release the oil causing the taste, which oil is not 
very easily oxidized. 
Naturally the cost of treatment with sulphate of copper will 
depend in part upon the amount of the chemical that is to be 
used, which in turn is determined by the kind of organism 
that it is intended to kill; but it may be said that a mixed growth 
of Mellosira and Asterionella was removed from the Troy reser- 
voir at an expenditure of 14.9 cents per million parts of water 
treated, labor included. The dose was one part of copper sul- 
phate to 3,500,000 parts of water by weight. 
In the article by Kellernan above quoted there are figures 
given indicating the safe limit for treating water with copper 
sulphate when certain fish are to be protected. 
COPPER SULPHATE, EXPRESSED AS PARTS, PER MILLION PARTS 
OF WATER 
Black bass 2.1 
Carp 3 
Catfish 4 
Goldfish 5 
Perch 75 
Pickerel 4 
Suckers 3 
Sunfish 1.2 
Trout 14 
It must be noted that these figures assume a thorough mix- 
ing of the sulphate solution with the whole body of water. They 
would not hold for the unequal distribution and resulting local 
concentrations already mentioned. 
Although copper sulphate is toxic to typhoid bacilli and has 
been proposed for water disinfection, yet the dose required is 
larger than that for the destruction of algae and its employment 
has not been uniformly satisfactory, in that it does not kill 
the pathogenic organisms with certainty. Moreover, for pro- 
tection against typhoid the administration of the dose would 
have to be continuous. For such work of disinfection better 
results can be secured by the use of bleaching-powder or liquid STORED WATER 
327 
chlorine. Simple storage of water for some hours in copper 
vessels will kill B. typhosus. In this connection it is interesting 
to note the following from Water, Aug. 15, 1905: 
"In the 'Ousruta Sanghita'-a collection of medical lore 
in Sanskrit, probable date 2000 b.c., chap, xlv., verse 15- 
appears this instruction: f It is good to keep water in copper 
vessels, to expose it to sunlight, and filter through charcoal.' 
" In the ' Neghrund Bhusan '-a collection of medical maxims 
from the ' Ayura Veda,' the earliest Sanskrit work on medicine 
extant, of about the same date-in the chapter on water, in the 
last sloka but two, it is directed to treat foul water by boiling 
and exposing to sunlight and by dipping seven times into it a 
piece of hot copper, then to filter and cool in an earthen vessel. 
It appears from the above that the principles of disinfec- 
tion and filtration were understood by the ancient inhabitants 
of India." 
In some reservoirs which have been formed by the extensive 
flooding of swamp bottoms there may develop objectionable 
growths of Crenothrix, a general term denoting a group of 
aquatic plants which at times give much trouble because of a 
tendency to develop in the street mains and clog the pipes. 
The growth is often discovered quite unexpectedly, being dis- 
lodged by the current attending hydrant flushing or by the 
draft caused by fire-engines. Dead ends are spots likely to 
harbor it, and its long rusty fila- 
ments have been mistaken for 
horse manure. 
There are three types of the 
growth, each possessing the pe- 
culiarity of precipitating from the 
water in which it grows its own 
particular metallic hydroxide. By 
far the commonest of the three 
is Crenothrix Kuhniana, which de- 
mands iron for its development and which deposits large 
amounts of iron hydroxide as the result of its growth. 
CRENOTHRIX. 328 
WATER-SUPPLY 
Crenothrix ochracea precipitates aluminum hydroxide and is 
yellow to yellowish white in appearance. Its usual color is that 
of ochre, as it nearly always precipitates some iron with the 
aluminum. The third species, Crenothrix manganijera, precipi- 
tates manganese hydroxide, and is dark brown or black in 
appearance. 
The growth of all three species is encouraged by a lack of 
oxygen, with a consequent reducing action in the soil or water. 
WATER PIPE NEARLY BLOCKED BY GROWTH OP CRENOTHRIX. 
Each of the three species seems to use a selective power 
in precipitating the special oxide ascribed to it, and about 
one-third of its dry weight is composed of the oxide selected, 
whether it be of iron, manganese or aluminum.* 
The iron required for growth must be in solution, and the 
quantity demanded would seem to be about 0.3 part of Fe per 
million. 
In order that the iron may be in solution, we naturally 
would expect the dissolved oxygen to be low and the quantity 
of reducing agents, such as organic materials, to be high, and 
those are the conditions that we find in practice to be favorable 
to the development of the plant. 
It is likely to be encountered in waters from swampy, peaty 
* See full.paper by D. D. Jackson, J. Soc., Chern. Ind., 1902: 681. STORED WATER 
329 
sources, where dissolved oxygen is scanty and where the neces- 
sary iron in solution may be had. Driven wells in such locali- 
ties. frequently furnish it. Darkness favors its growth, and its 
development in city water-mains is often excessive, resulting 
in a material reduction of the carrying capacity of the pipes. 
The writer has some doubt about the " manganese " variety 
of crenothrix being as rare as some think it is, he having found 
large quantities of manganese in a heavy Wisconsin growth. 
Beythien and others have, moreover, noted that the presence 
of manganese in water directly favors the growth of the ordi- 
nary form of crenothrix. 
Beyond the mechanical stopping of street pipes, crenothrix 
is exceedingly objectionable to the laundry interests of the com- 
munity, for the reason that its rusty filaments cause 11 iron 
stains " to appear upon white linen. 
Removal of the iron by oxidation and filtration is the best 
guard against troubles due to crenothrix. 
The stripping of reservoir sites and the removal, so far as 
possible, of all the vegetable material of the upper soil, has been 
advocated and carried into practice with a view to diminish 
the amount of plant food and thereby avoid the troubles arising 
from the excessive growth of the taste- and odor-producing 
organisms. 
Decomposition of recently killed vegetation takes place 
under water quite rapidly at first, but the process is shortly 
converted into one of exceeding slowness, particularly where 
the covering water is deep. So permanent, in fact, is timber 
which has been deeply submerged that the oaken piles which 
in prehistoric times supported the buildings of the Swiss " lake- 
dwellers " are still firm and solid, although black in color. On 
the other hand, alternate flooding and exposure to sun and air 
is quickly destructive of vegetable matter, and as a result a 
reservoir with very gentle sloping sides furnishes conditions 
less favorable to a high-grade water-supply, particularly if 
there be considerable variation in the water-level. Even though 
the high-water mark be always maintained, sloping sides per- 330 
WATER-SUPPLY 
mit thin layers of water to be heated by the summer sun, 
thus encouraging abundant growth of aquatic plants, if there 
be sufficient food present for their nourishment. These plants 
subsequently decay to the damage of the water. 
It is especially undesirable to have the bottom of a storage- 
reservoir remain exposed for more than one season, for the 
reason that vegetation may develop in such quantity as to 
injure the water when the bare slopes are again submerged. 
The Boston authorities removed the soil from the site of the 
Wachusett reservoir (6.46 square miles or 4135 acres in area) 
LAKE VYRNWY (LIVERPOOL WATER SUPPLY). 
at a cost of $2,536,000. Some 800 holes were dug to determine 
the average depth of the top-soil layer, and it was concluded 
to take away 9 inches from the wooded portions and nj inches 
from the cleared lands.* 
Such stripping cost about thirty-six cents per cubic yard 
but this did not include tree and bush removal. 
As a provisional standard, i| to 2 per cent of organic matter, 
as determined by the loss of ignition of a sample of earth dried 
at ioo° C. (2120 F.) was fixed upon as the permissible limit 
of organic matter that might be allowed to remain on the 
bottom and sides of the reservoir. 
No stripping of the soil from the bottom of the Vyrnwy 
* Rep. Mass. Board of Health, 1895. STORED WATER 
331 
reservoir, supplying Liverpool, nor from the Ashokan reser- 
voir lately completed for New York, was done. 
At Columbus, 0., the reservoir built to store the water of the 
Scioto river has an area of about 255 acres. 
" It was not deemed necessary to remove any of the surface 
soil except to a very limited extent around former human 
habitations within the submerged area; but responsive to the 
requirements of the Ohio State Board of Health, all vegetable 
growths were cut down even with the ground, being then gathered 
into heaps and burned, while the stumps and roots of all trees 
and shrubs one inch or more in diameter were grubbed to a 
depth of one foot below the surface and similarly gathered and 
burned. 
DETAILED COST OF CLEANING AND GRUBBING RESERVOIR SITE 
COLUMBUS, OHIO 
Item. 
Days. 
Rate. 
"Total. 
Per Cent, 
of Cost. 
Superintendent 
25S 
$4,165 
$1,063 
2.6 
Time keepers 
255 
1-75 
446 
I. I 
Foremen 
1,030 
2.50 
3,325 
8.2 
Foremen 
205 
2.00 
410 
1.0 
Carpenter 
54 
2.00 
108 
■3 
Dynamite men 
435 
i-75 
761 
1.9 
Laborers 
i4,49i 
1-5° 
2i,737 
53-3 
Single horse 
222 
i-5° 
333 
.8 
Two-horse team 
847 
3-5° 
2,964 
7-3 
Dynamite 
68,000 lbs. 
7,820 
19.1 
Machinery and repairs 
Total 
1,800 
4-i 
Cost per acre, $159.50. 
A portion of the bottom land of about 36 acres had been 
cultivated but the remainder was thickly covered with a growth 
of trees and shrubs common to the forests or groves of the 
locality, some of them being five feet in diameter. The bot- 
tom lands were very fertile, and everywhere, in addition to the 
trees supported rank weeds, growing to a height of eight to 
thirteen feet." * 
* See paper by Julian Griggs, read before the annual convention of the Amer- 
ican Society of Municipal Improvements, Birmingham, Ala., October, 1906. 332 
WATER-SUPPLY 
The stripping of reservoir sites and the removal of a portion 
of the upper soil entails very great expense when the surface 
to be stripped is at all extensive, as in the instance of the Wa- 
chusett reservoir supplying Boston. 
In their report upon the probable cost of stripping the sur- 
face soil from the Ashokan reservoir site, Messrs. Hazen and 
Fuller stated it would possibly reach the great figure of 
$5,000,000. 
In view of the expense of such treatment for large reservoirs, 
the question is pertinent, " Does it pay? " 
At Holyoke, Mass., the annual water report for 1908 says: 
11 Great care had been taken in cleaning and stripping the reser- 
voir by removing all vegetable and organic matter, thus lessen- 
ing to a minimum the food supply for supporting living organ- 
isms in the water. The thorough cleaning of the reservoir has 
not been wholly successful, as an aquatic plant known as 
' Chara ' has grown and flourished in the reservoir all summer 
and imparted to the water a taste and odor that made it unfit 
for drinking or even for cooking purposes." 
Mr. J. M. Diven * has had interesting and contrasting 
experiences with both stripped and unstripped reservoirs: 
11 The Elmira reservoir was as thoroughly stripped as pos- 
sible; great care was taken to keep out the first washing from 
the drainage area and the muddy flood waters. There was little 
or no marsh land on the drainage area, the catchment area being 
seemingly ideal. The reservoir was clean and clear; on the 
sides the slopes were abrupt, and there was very little shallow 
water. 
" At Charleston, S. C., the drainage area was largely swamp, 
and there was much decayed vegetable matter on all of the area 
drained, the water being decidedly peaty. The reservoir 
covered a large surface, was shallow, and absolutely unstripped 
or even cleared. Much of the land flooded was composed of 
black muck or decayed vegetable matter. 
" In the first case (Elmira) the conditions were at the first 
satisfactory and the water good for several years. But trouble 
* American Water-works Asso., 1908. STORED WATER 
333 
from algal growth came in time and has steadily grown worse, 
in spite of strenuous efforts to remedy the condition. 
" The second case (Charleston) was troublesome and un- 
satisfactory from the first, but has somewhat improved and 
promises to continue to improve." 
On the other hand, Wachusett reservoir, above referred to, 
has given satisfaction from the first. The shores are clean 
and the water clear and attractive. No odors have developed 
and the color of the entering water has fallen from 45 to 18 
through the bleaching action of open storage. All B. coli 
are removed by the passage through the reservoir. One 
evidence of the beneficial result of stripping in this instance is 
the fact that the bottom water, 102 feet down, is as good or 
slightly better than that of the surface; at least such was its 
appearance at the time of the writer's visit. Inasmuch as this 
reservoir is the best illustration that we have of " stripping " 
on a large scale, it will be well to add here the analytical results 
of the reservoir water and of those of the entering streams.* 
CHEMICAL ANALYSES-WACHUSETT RESERVOIR 
(Parts per Million) 
July 7, 1914 
Quinnepoxet 
Inlet. 
Stillwater 
Inlet. 
Surface near 
Dam. 
Bottom 
near Dam. 
Turbidity 
V. slight 
Slight 
V. slight 
V. slight 
Sediment 
Slight 
Considerable 
V. slight 
V. slight 
Color 
45- 
34- 
18. 
20. 
Odor, cold 
F. veg. 
F. veg. 
None 
V. F. veg. 
Odor, hot 
Dis. veg. 
Dis. veg. 
V. F. veg. 
F. veg. 
Residue on Evaporation: 
Total 
43-5 
53-5 
32-5 
38.5 
Loss on Ignition 
14-5 
16.0 
9-5 
10.0 
Fixed 
29.0 
37-5 
23.0 
28.5 
Free ammonia 
0.024 
0.032 
0.022 
0.046 
Albuminoid Ammonia: 
Total ' 
0.284 
0.264 
0.170 
0.130 
Solution 
0.204 
0.176 
0.126 
0.126 
Suspension 
0.080 
0.088 
0.044 
0.004 
Nitrates 
none 
none . 
0.010 
0.020 
Nitrites 
O.OOI 
O.OOI 
none 
none 
Oxygen consumed 
5-9 
4-8 
2-4 
2-4 
Hardness 
5- 
14- 
10. 
8. 
Iron 
2.2 
o-37 
0.03 
0.01 
* Analysis by State Board of Health 334 
WATER-SUPPLY 
More recently R. H. Stearns * reports the color reduction 
of water in the Wachusett reservoir, to be 66 per cent following 
a storage period of 615 days. The data show the color of the 
influent water to be 43.5 parts per million and that of the 
effluent 14.7 parts per million. 
While the latter figure is readily secured it is not so easy to 
obtain a true value for the former, because of the many differ- 
ent waters that may flow into a large reservoir; nevertheless 
there is manifestly a decided bleaching effect to be credited 
to the action of the prolonged storage. 
Such a result of stripping as is illustrated by the Wachusett 
reservoir is not to be expected always. The Boston Metropoli- 
tan water supply is caught in a granite region and the streams 
furnish waters which, though colored, are not turbid. The 
water flowing into Wachusett reservoir is low in plant food 
and, therefore, as all such food was stripped from the reservoir 
bottom during construction, there is nothing to maintain 
heavy algal growth. Should a reservoir of similar size be 
constructed elsewhere, say in the Middle West, or wherever 
turbid stream flow is common, then so good a return for 
the money expended in the stripping could not be secured 
and the stripping would not pay. This is but another in- 
stance of how desirable it is to judge of each case upon its 
own meritsr 
The writer advocates the expenditure of comparatively 
little money in the preparation of sites for large storage-reser- 
voirs, except under the conditions above noted, for the reason 
that, although thorough stripping will likely give immunity 
from algal growths for some years, yet freedom from the occur- 
rence of taste and odor in the stored water may not last for 
long. Sooner or later there will probably be carried into even 
the most carefully cleaned reservoir enough food material to 
sustain a plankton growth of a density governed by the local 
conditions. 
The character of the tributaries must be considered as well 
as the nature of the bottom of a proposed reservoir, for it is 
* J. N. E. Water-works Asso., March, 1916. STORED WATER 
335 
manifestly loss of money to improve the latter if the former can 
quickly replace much of what has been taken away. 
For the sake of general appearances, if for no other reason,, 
trees, shrubs and bushes should be removed. Dead, standing 
timber and fallen logs are most unsightly and are very likely 
to produce complaint from the visiting public. In other words, 
the reservoir site should be cleared and grubbed, with, of course, 
entire removal of every vestage of human habitation; but 
beyond that it usually does not pay to go. The portion of the 
flooded land lying between high-water and low-water marks 
should receive especial attention, for the reason that during the 
periods of it's exposure it is capable, if uncared for, of presenting 
an unpleasant appearance and provoking adverse criticism; 
with the further objection that heavy weed growth may develop 
if it be long uncovered, which growth will contribute toward 
the production of taste and smell when the water again 
covers it. 
Broadly speaking, an " old bottom " is better than a new 
one, because it is likely to contain less plant food; but the 
rule has many exceptions. 
Even natural lakes are frequently seen " in bloom,"-that 
is, loaded with minute life,-and they so remain for a period 
during which their waters are not acceptable for domestic 
use. 
Aeration, filtration, and the judicious, occasional use of 
copper sulphate constitute the processes at our disposal for 
combating the annoyance arising from algal growths, and 
their use will ordinarily give greater satisfaction than the 
expensive stripping of reservoir bottoms. 
Successful removal of very considerable odor due to growth 
of algae has been secured by the use of double sand filtration 
and double aeration at South Norwalk, Conn. A detailed 
description of the plant is given in J. N. E. Water-works Asso., 
for March, 1916 by H. W. Clark, through whose advice the work 
was undertaken. The good results obtained are similar to 
those following double filtration at Springfield, Mass., by a 
plant whose operation is described as removing " the organ- 336 
WATER-SUPPLY 
isms and the tastes and odors and causing a very substantial 
reduction of the iron, organi cmatter and color of the water 
treated." 
Depth of reservoirs is not so important as the presence 
of food-supply in the matter of the existence or absence of 
organisms. The Massachusetts Board of Health reports the 
case of Pilling's Pond, a very old storage-reservoir, eighty-five 
acres in surface, with an average depth of only three feet. No 
abnormal growths appeared in this reservoir, not did the water 
become offensive although its temperature at times reached 
8o° F. The explanation offered is that, owing to the age of 
the reservoir, the bottom mud no longer contains food-supply/1'' 
The deepening of shallow flowage near shore can be best accorm 
plished by the method recommended by FitzGerald and sketched 
herewith. 
WATER LINE - 
OLD BOTTOM- 
NEW BOTTOM .. 
It is particularly desirable that no depressions should be 
permitted to remain upon a reservoir bottom, which could 
be cut off from the receding water as the basin becomes drawn 
down, and thereby form stagnant pools. Such spots would 
surely prove breeding-places for algal growth. Free drainage 
should be provided for all pockets of this kind, or, better still, 
they should be filled up. It may be said here that the work 
of stripping a reservoir bottom is sometimes complicated by 
the occurrence of deposits of soft material rich in organic 
products and too deep to be economically removed. It is 
good practice to fill in the surface of such places with good 
* Rep. Mass. Board of Health, 1890 [1], p. 749. STORED WATER 
337 
sand or gravel, after having taken away the more superficial 
portion. 
Such filling need be but a few inches thick in order to pro- 
tect the water from damaging contact with the bed of muck 
below, providing no upward movement of ground-water, as 
from a spring, is taking place, in which case even a much thicker 
layer of gravel would be of no avail. 
Sulphuretted hydrogen may add its disagreeable smell to 
the odors occurring in new reservoirs, particularly shallow ones. 
The decomposition of vegetable material, killed by flooding, 
causes a reduction of the sulphates present to sulphides, and these 
sulphides are further acted upon by the acids also formed by 
such decomposition, with liberation of the foul-smelling gas. 
The author found this gas on one occasion due to a some- 
what unusual cause. The reservoir-dam had been built 
of blast furnace cinder, and the water was, in consequence, 
strongly impregnated from the sulphur compounds contained 
therein. 
Waters from underground sources should be distributed for 
use as soon as possible after they have been brought to the 
surface; for, as we have seen, they are commonly well supplied 
with plant-food in solution, and, under the influence of sun- 
light, there is danger of abundant development of objection- 
able algae if much time for open storage be allowed. 
With surface-waters the case is quite the reverse, and long 
storage becomes a distinct advantage. 
Sedimentation of suspended matter, and destruction of 
bacteria by simple lapse of time, are two sources of benefit 
arising from the impounding of surface-water. 
Bacteria often die but slowly, and although a large per- 
centage of their number will perish through storage, it should 
not be forgotten that they are very small and very light, 
and consequently are very long in settling, expecially in clear 
water; so that it should not be expected that a reservoir could 
do the efficient work accomplished by a filter. 338 
WATER-SUPPLY 
Sedimentation of small particles is slower in winter than 
in summer, owing to the increased viscosity of the colder 
water. 
The water for Cincinnati was taken from the Ohio River 
and delivered to the consumers with but a few hours' subsidence 
in the Eden Park reservoir, while the water of Covington was 
carried for an average of thirty-two days in the subsiding- 
reservoirs before it was delivered to the people. 
From these results the average reduction of bacteria in the 
Ohio River water by thirty-two days' subsidence is shown 
to have amounted to nearly 94 per cent. 
The following table, showing the influence of sedimentation 
as measured by the number of bacteria per cubic centimetre of 
water, is taken from the report of the Cincinnati Engineer 
Commission: 
Bacteria, per c.c. of Water. 
Reduction of 
Bacteria by Sedi- 
mentation. 
Cincinnati 
(Raw). 
Covington 
(Settled). 
1472 
272 
88.32 
1599 
194 
87-87 
5062 
172 
96.60 
- 
182 
96-4I 
1656 
53 
96.80 
2042 
56 
97.26 
1561 
63 
95-96 
1526 
75 
95-°9 
684 
20 
97-08 
329 
26 
92.10 
1232 
112 
90.91 
1144 
84 
92.66 
1436 
102 
92.90 
Cincinnati has now a new filter plant; but the above fig- 
ures, applying to past conditions, still illustrate as fully as ever 
the value of sedimentation. 
Percy Frankland found the following numbers of bacteria 
per cubic centimetre in Thames water at the intake of the 
Grand Junction Company, and in water from the large 
reservoir of the company, where the greater part of it STORED WATER 
339 
had been stored for six months, and none for less than one 
month: 
Intake 1991 bacteria 
Reservoir 368 " 
The West Middlesex Company causes its water to pass 
through two storage-reservoirs before it is delivered upon the 
filter-beds. The influence of such passage is seen in the fol- 
lowing counts of bacteria as made by Frankland: 
Intake at Hampton 1437 per cubic centimetre 
After passing first reservoir 318 
After passing second reservoir 177 
The value of sedimentation was shown at Philadelphia 
during the prevalence of typhoid fever in that city in 1891. 
"By much the highest mortality is in the twenty-ninth and 
thirty-second wards. This is an elevated section of the city, 
newly improved and occupied for the most part by well-to-do 
people. The drainage is good and the laws of health are 
doubtless as well observed as in any other portion of the city. 
But these wards are too high to draw water from the subsiding- 
reservoir, and they are accordingly furnished by direct pump- 
age from the river. This is the case also in the twenty-eighth 
ward adjoining, and the district so supplied extends southward 
including the fifteenth ward, another well-to-do part of the 
city where typhoid is especially prevalent. These four wards 
furnished by direct pumpage, have a population of 184,000, 
and report 317 cases of typhoid fever, or at the rate of 172 
to 100,000 inhabitants." * 
The favorable influence of precipitating mud in hastening 
the fall of bacteria was demonstrated in this laboratory some 
years since. The result was to have been expected in view of 
the well-known tendency of falling solids to drag down other 
matters with them. Similar results were observed by Kruger, f 
* See also " Bacterial Reduction by Storage," Engineering Record, Aug. 13. 
1898; " Influence of Dredging Lake Bottom, Public Water-supply," Hill, p. 96, 
t Zeit. f. Hygiene, VII. 86. 340 
WATER-SUPPLY 
Dr. A. C. Houston, of the London Metropolitan Water 
Board, has undertaken some very extended researches upon the 
question of water purification as a result of storage. He found 
that in stored Thames water the death of typhoid bacteria took 
place rapidly, although the rate varied with the temperature 
of the water. In cold water they lived longer than in warm, 
and 500 F. seemed to be a critical point above which their 
mortality rate was much increased. 
In his 7th Research Report, Houston states that typhoid 
bacilli lived in stored raw Thames water for the following 
lengths of time: At 320 F., five weeks; at 410 F., four weeks; 
at 500 F., three weeks; at 64° F., two weeks. 
Even these figures do not tell the entire story. Put in more 
detail they read: 
At start. 
1 week. 
2 weeks. 
3 weeks. 
4 weeks. 
5 weeks. 
32° F.............. 
103,328 
47,766 
980 
65 
34 
3 
41° F 
103,328 
14,894 
26 
6 
3 
- 
5o° F 
103,328 
69 
14 
3 
- 
- 
64° F 
103,328 
39 
3 
- 
- 
- 
He concludes: 11 It is difficult to escape the belief that thirty 
days' storage of river-water is tantamount to sterilization, so 
far as the microbes associated with water-borne epidemic disease 
are concerned." 
When experimenting with an artificially infected water to 
determine the effect of storage upon the typhoid bacillus, Dr. 
Houston felt that any error so introduced was upon the side of 
safety, ' because he had previously shown the "cultivated" 
typhoid organism to have a greater longevity than the " natural " 
Bacillus typhosus. In his report he dwells at length upon the 
advantages to be derived from " adequately storing the raw 
impure river waters." Even if there were no economic reason 
for storing a river-water before rather than after filtration, yet 
it would be well to follow that course, aside from any question STORED WATER 
341 
of agal growths, for the reason that sedimenting silt greatly 
assists in bacterial removal. 
Dr. Houston further says: " I am well satisfied that a well- 
stored, rapidly filtered water is likely to be safer than an 
unstored, slowly filtered water." 
It is possible to go even further than this, for one can 
see how dangerous it might be to deliver, directly to the 
consumers, the water of a small and apparently pure moun- 
tain stream. The dejecta of a single typhoid patient would 
render so small a volume of water highly infectious, if no 
storage intervened, and an outbreak of the disease might 
follow. 
Although Dr. Houston is doubtless sound in his judgment 
that a great measure of safety will result from four weeks' 
reservoir storage of a polluted water, yet we must be assured 
that the period of storage is real and not simply apparent; 
or, in other words, we must know that all of the water really 
does remain in the reservoir for the specified length of time 
before it is used for public consumption. 
Where the inlet and outlet of a reservoir are near together, 
as is not uncommonly the case, it makes but little difference 
what the capacity of the total storage may be; the water 
simply slips in and out again with practically as little stay as 
though the reservoir were a stand-pipe. 
If the lake be long, narrow, and deep, and all of its water be 
obliged to traverse its entire length before being taken for sup- 
ply, then the conditions would appear ideal for purification of 
the inflowing water before the outlet was reached, and yet even 
under those excellent conditions it is possible to have introduced 
unexpected and upsetting factors, as is instanced by the history 
of the typhoid epidemic at Auburn, N. Y. 
Lake Owasco is one of the so-called " finger lakes " of 
western New York. Its length is about ten miles, breadth one 
mile; its watershed is about 190 square miles; and its depth 
is about 175 feet. A small stream enters its head, and Auburn, 
a city of some 30,000 inhabitants, has an intake, located at the 
north end or foot of the lake and forty feet below the surface. 342 
WATER-SUPPLY 
The temperature of the water at that point in May, 1913, I 
found to be 420 F. 
The peculiar feature of the case which has special interest 
here is the possibility of polluting material of fecal character 
being transported from a village near the head of the lake, 
down the inlet stream, and then northward for the entire ten 
miles of the lake's length to the Auburn intake situated near 
the lake outlet. 
We have all faithfully held to the dictum that " sedimenta- 
tion and time " are the great purifying agencies upon which to 
rely for the natural improvement of a once polluted water; 
and it takes a good deal of evidence to persuade us that sewage 
of a small village could make the trip down such a lake in a length 
of time and in such a manner as to dangerously affect the water of 
the lower end. Experimental data, however, have been secured 
showing that such a result can actually take place. Investiga- 
tion showed the following facts: The village sewage was, of 
course, small in volume, but during the winter months it was 
deposited at several points upon the banks of the inlet stream 
and there it collected in a more or less frozen condition until 
the occurrence of the spring thaw, at which time there was oppor- 
tunity for much accumulated fecal material to be washed into 
the lake in a state of suspension. There was also a chance 
of its being actually ferried upon cakes of ice, for the reason 
that certain privies were located upon bridges and fecal 
matter was dropped upon the very centre of the ice-covered 
stream. 
As stated, the shape of the lake is long and narrow and its 
axis lies north and south. It must be further noted that the 
prevailing wind is from the south, with a tendency to blow 
the surface-water directly toward the city intake at the north 
end. 
By means of triangulation and the use of floats constructed 
so as to be moved by water currents existing at the different 
depths of from five to twenty feet, it was ascertained that the 
upper stratum of water moved northward with the wind, as 
would have been expected. The rate of this movement being STORED WATER 
343 
ascertained, it was found that with relatively light winds the 
movement of the water down to a depth of five feet amounted 
to about 3 per cent of the wind movement, while at lower depths 
this water movement diminished to as low as three-quarters of 
one per cent of the wind movement. Thus, to quote from the 
figures of Mr. Ackerman, who made these tests, with a wind 
movement of six miles per hour the percentage which the 
water movement was of the wind movement was as follows: 
At 5 feet depth, 3.2 per /cent; at 10 feet depth, 1.74 per 
cent; at 15 feet depth, 0.87 per cent; at 20 feet depth, 0.75 
per cent. 
With a higher wind velocity the water also travelled with 
greater velocity, but its movement was then not so large a 
percentage of the wind movement. Thus, with a wind blowing 
17 miles an hour the water movement at a depth of five feet 
amounted to but per cent of that of the wind. From these 
data it was easy to calculate that pollution entering the head of 
the lake could make the trip to the foot of the lake in three days 
or less. 
Knowing, as we do, from Dr. Houston's experiments, that 
cold water below 500 F. will favor the longevity of the typhoid 
fever bacillus, it is easy to see how entirely possible it would 
be for living germs to reach the intake in dangerous con- 
dition. 
It will be noted how striking is the resemblance which some 
features of this case bear to the classic instance of the outbreak 
of typhoid fever at Plymouth, Pa., where the whole trouble 
came from dejecta of a single individual being thrown out 
upon a hillside where it froze and accumulated for weeks and 
finally, upon the coming of the thaw of spring, was washed into 
a stream tributary to the city reservoir. This sudden washing 
of accumulated fecal material furnished in both of these instances 
a volume of pollution out of all proportion to the amount which 
would be daily derived from the contributing population dur- 
ing ordinary times of fair weather, and as a result it overtaxed 
and broke down nature's ordinary means of purification and 
protection. 344 
WATER-SUPPLY 
There is no question but that this particular case, showing, 
as it does, the dangers that may arise from such winter accumu- 
lation, and showing further the possibility, under favorable con- 
ditions, of the transportation of such material over considerable 
distances in a lake, will cause many of us to materially amend 
our notions about the dependence to be placed upon lake and 
reservoir storage as a means of protection against the evils 
following water pollution. We should not trust to simple 
storage without a thorough knowledge of just how it is being 
accomplished. The writer has in mind an instance of a large 
lake some five miles in length which has a stream entering 
within one mile of a city intake, and, because of the entering 
water having a low specific gravity, there is a possibility of its 
flowing over the surface of the lake toward the intake whenever 
the wind is in the right direction. The great length of that lake 
is, under such circumstances, of small value for purification 
purposes. In another instance a cold stream entered the 
reservoir and its water hugged the bottom until the intake was 
reached, hence the low level gate drew unstored water. 
All of this certainly goes to show that we should be cautious 
about banking too strongly upon the efficiency of reservoir 
purification under all circumstances, and it demonstrates the 
necessity of our being well acquainted with the conditions 
surrounding each individual case before venturing an opinion 
on the matter. 
It should be noted here that, in judging of the bacterial 
efficiency of lake or reservoir storage, the interpretation of the 
results of an examination may be obscured by an increase 
in the total count of bacteria reported due to the disturbing 
influence of the spring or autumn " turnover." 
Doubtless the most apparent advantage to be obtained from 
storage is the removal, by sedimentation, of those suspended 
materials, mostly of mineral character, which cause unsightly 
turbidity in water. Quiescence is naturally contributory to 
such removal while " scouring velocities " tend toward an 
opposite result. STORED WATER 
345 
SETTLING-TANKS SHOWING " SKIMMING " DISKS. DENVER, COLO. 
SAME TANKS EMPTY OF WATER SHOWING DEPOSIT OF SEDIMENT. 346 
WATER-SUPPLY 
Dappert gives the velocities required to move sundry mate- 
rials as follows.* 
Ft. per Second. 
Very fine clay 0.40 
Very fine sand 0.70 
Light vegetable soils 0.83 
Sand as coarse as flax seed 0.91 
Gravel, size of a pea . 1.40 
Small pebbles 2.52 
Pebbles one inch in diameter 3.00 
Pebbles size of hen's eggs 4.15 
The theory of the clearing of water by settlement, and the 
most economic size for purely sedimentation-reservoirs, are 
questions which have been exhaustively discussed in our engi- 
neering societies by such men as J. A. Seddon, Whitney, and 
others, and reference must be had to their voluminous -papers 
for full information. 
But it may be here said that the bulk of sediment capable 
of falling within a reasonable length of time will be found to 
deposit within about twenty-four hours, the balance settling 
out with much slowness. It is therefore very questionable 
economy to provide for a sedimentation period of over thirty- 
six hours. 
Percentage of suspended matter removed from water by 
twenty-four hours' plain subsidence.! 
Kansas City, Missouri River 82 per cent 
Cincinnati, Ohio 62 11 
New Orleans, Mississippi 45 11 
Whitney calls attention to the fact that " as the material 
in suspension grows finer the weight of each particle decreases 
so much more rapidly in proportion than its surface that 
there is, relatively, a larger amount of surface area in these 
fine particles, and a great deal of surface friction in their 
* Engineering News, 55: 126. 
f J. N. E. Water-works, Asso., xvii. 165. STORED WATER 
347 
THE CAPEDIMONTE RESERVOIRS, NAPLES, ITALY. 348 
WATER-SUPPLY 
movement through a medium. Consequently they settle very 
slowly." 
" Ordinary convection currents, induced by normal changes 
of temperature, would be sufficient to keep these fine particles 
in suspension, as it is known that currents of air keep fine 
particles of dust and ashes in suspension." * 
One of the most peculiar forms of settling-chamber the 
author has seen was prepared for use by the Denver Union 
Water Company, at Denver, Colorado. It is readily under- 
stood by reference to the illustration on page 345. The tanks 
are each 20 by 20 by 3I feet, and water is admitted through nine 
rectangular, orifices near the bottom. The effluent is. 
" skimmed " from the surface by ninety 16-inch disks which 
connect with the underdrains by means of one-inch vertical 
iron pipes. The 15-inch partitions, shown on the bottom, 
are arranged to assist in the flushing out of the sediment. At 
the time of the writer's visit the efficiency of the settling-tanks 
was represented by a removal of a large part of the suspended 
matter and of 32 per cent of the bacteria present in the raw 
water. 
We have already seen that exposure of subterranean waters 
to sunlight commonly results in development of objection- 
able vegetable growths. The somewhat peculiar reservoirs 
at Paris which hold the Vanne spring-water, supplying a large 
portion of the drinking-water used in the French capital fur- 
nished an illustration of dark storage. The springs are about 
107 miles distant, and the grade of the conduit-pipes is 1 centi- 
metre per 100 metres (0.4 inch per 328 feet). In order to 
secure sufficient storage, and yet to economize space within the 
walls of Paris, two reservoirs were built, one on top of the 
other.f The lower one is constructed of concrete, and 1800 
concrete columns support the upper story, which is of brick 
This upper chamber is covered by a roof which rests on brick 
continuations of the concrete columns. The water area in 
* Wiley's " Agric. Anal.," i. 180. 
t The Montmartre reservoirs of Paris of three superimposed chambers instead 
of two. STORED WATER 
349 
each reservoir is 272 by 136 metres (892 by 446 feet), and its 
depth in the upper one is 8| feet and in the lower one 131 feet. 
The total storage capacity is 200,000 cubic metres (52,800,000 
U. S. gallons). The temperature of the water is constantly 
48.2° F. in winter and 51.8° F. in summer. No trouble has 
ever been experienced with algal growths or odors. Cleaning 
takes place but once in five years, at which time about half an 
inch of compact hard deposit is removed. The reservoirs hold 
a supply sufficient for about six days. 
Another underground storage system which will repay a 
visit is that of Naples, Italy. Old quarries lying in the hill of 
Capedimonte overlooking the city have been enlarged and 
cemented with the result shown in the illustration on page 
347. Five parallel tunnels have thus been formed some 150 
feet below the surface of the ground. Their dimensions are 
35.4 feet in height and 30.3 feet greatest width. The depth 
of water is 26 feet, and the galleries are separated by a space 
of rock left unexcavated. The capacity of the five basins is 
21,120,000 U. S. gallons. 
Underground storage of water is also to be found at Gib- 
raltar, and is peculiar in the respect that military precautions 
are considered in its accomplishment. The writer having had 
special opportunities for examining this particular system of 
water-works, a short description of what he saw may not be 
out of place. 
Gibraltar is, beyond all things else, a fortress. The visitor 
is continually reminded of this fact, not only by the visible 
signs of military occupation everywhere present, but also by 
the rules and regulations with which the place positively 
bristles. The assumption upon which the garrison appears to 
act is that a state of war continually obtains, and that a hostile 
attack should be momentarily expected. 
So carefully are the military secrets of the rock guarded 
that even the officers of the post are individually not fully 
informed as to the nature of the defences outside of the sections 
where their duties lie. To such an extent is this carried that 
the engineer officer whose duty it was to design the works for 350 
WATER-SUPPLY 
"NORTH FRONT," GIBRALTAR. STORED WATER 
351 
the new water-supply told the writer that he could not go up 
the west face of the rock until his pass had been signed in 
England. 
It may be easily imagined that the foresight which goes so 
far as to cause the garrison to be fed on " siege-rations " one 
day in each week would not permit the possibility of such 
garrison being deprived of its water-supply through the opera- 
tion of an enemy. 
The area of " Gibraltar " is small, but that portion of it 
which is within the old walls and which constitutes the " city " 
is densely populated. Roughly speaking, the civil popula- 
tion may be placed at 20,000, while the garrison numbers 6000. 
These figures would be, of course, liable to modification in 
time of actual war. 
The bold-faced " Rock of Gibraltar," with the pictures of 
which we are all so familiar, rises sheer from the " Devil's 
Tower Road," on the " North Front," and faces the land side, 
not the sea as most people fancy. It looks toward Spain, 
and makes the approach of an enemy from that direction all 
the more difficult. The east side is also very precipitous; 
while the west slope, although decidedly steep, permits of 
enough fairly level ground being found at its base to furnish 
a foothold for the " city " with its circumvallating walls. 
Gibraltar is not limited to a single source of water-supply. 
For ordinary uses, such as flushing, street-sprinkling, and the 
like, a brackish water is used which is derived from wells sunk 
in the flats north of the rock, and near the neutral grounds. 
The supply so secured is about 25 per cent sea-water in winter 
and about three times that amount in summer, owing to the 
variation in seasonal rainfall. It is pumped to a tank located 
in the " Moorish Castle," a portion of the ancient defences, 
and flows thence to the town by gravity. This water is 
metered and charged for at the rate of | peseta for 100 Imperial 
gallons, or about 40 cents per 1000 U. S. gallons. 
Situated not far from the " Alameda Gardens," on the 
west slope of the rock, is a distilling-plant for furnishing a 
potable supply from sea-water. The apparatus is well pro- 352 
WATER-SUPPLY 
tected and ready for an emergency, but is not ordinarily in 
use. 
The principal water-supply for Gibraltar is obtained from 
CROSS SECTION 
Longitudinal section. 
SECTIONS OF STORAGE-TUNNEL RESERVOIRS AT GIBRALTAR. 
SKETCH-SECTIONS OF NEW STORAGE-TUNNEL RESERVOIR AT GIBRALTAR. 
the rain, and a careful inspection of the west face of the rock 
will show a whitish patch near the top which acts as the catch- 
ment area. This area contains about sixteen acres, and has 
been carefully cemented at a cost of i| pesetas (about 30 cents) 
per square yard. The water falling thereupon, averaging 33I STORED WATER 
353 
WEST FACE OF ROCK, GIBRALTAR (SHOWING CATCHMENT AREA IN UPPER LEFT CORNER). 354 
WATER-SUPPLY 
inches annually, passes through a rough strainer of " polar- 
ite " and sand, and is thence delivered into the underground 
storage-reservoirs. 
The catchment area above described has been recently 
much extended and embraces a portion of the east face of the 
rock. 
The original four storage-reservoirs are of the same size, 
and are of tunnel form, being 201 feet long, 22 feet wide at the 
top; and 18 feet wide at the bottom, with a height of 51 feet, 
see cross-section, page 352. The sides are stepped off by 4-inch 
ledges every 7 feet, measured vertically. The sides, bottoms, 
and ends are concreted and then plastered with four coats of 
a mixture of equal parts of cement and fine limestone. The 
filtered rain-water is admitted to each tunnel through an 18-inch 
cast-iron pipe, carried to the bottom of the chamber. 
Security against the shell-fire of an enemy is not the only 
advantage of this form of water-storage, for it must be remem- 
bered that the absolute darkness which prevails under such 
conditions inhibits the growth of aquatic vegetation, and the 
temperature of the water is maintained constantly at a low 
degree, in this instance 500 F. 
The exact location of these reservoirs is not permitted 
the outsider to know, although their general form and dimen- 
sions are as given above. The internal measurements are 
correct, as stated, but the true angle of slope of the exterior 
rock-face is intentionally withheld. It will be noticed that the 
exterior ends of the tunnels are not excavated. This is for the 
protection of the water-supply against the artillery-fire of a 
besieging force. 
A newer and much larger storage system has been driven 
through the entire rock from side to side, excepting, of course, 
the unexcavated ends. Exact information covering this de- 
velopment it is, naturally, impossible to secure, but the rough 
sketch given and the quotation here following will sufficiently 
illustrate its general character. 
"The annual rainfall during the last 114 years has fluctu- 
ated between 77 inches (1855-6) and 15 inches (1800-1), and STORED WATER 
355 
COVERED RESERVOIR. PASADENA. CAL. 356 
WATER-SUPPLY 
during that time there have been twenty-three years with more 
than 40 inches of rainfall and twenty-two years with less than 
25 inches. In order to provide water by pipes to the whole 
of the Fortress in the event of three successive years of mini- 
mum rainfall, it will be necessary to have a collecting area of 
37 acres and storage reservoirs of 17,000,000 gallons capacity. 
A collecting area of 10 acres has been provided on the sand 
slopes above Catalan Bay. This area is capable of yielding 
240,000 gallons per inch of rainfall, i.e., an increased yield 
during average rainfall of 8,000,000 Imperial gallons per annum. 
The sand has been covered with corrugated iron sheets laid on 
a creosQted timber framing. With ordinary care this covering 
will be as durable and efficient as the older areas, and much 
less absorbent. The water from the area is collected in a 
channel at the bottom of the slope and conveyed in a tunnel 
about 2030 feet long through the rock to the four reservoirs 
constructed four or five years ago. The channel, both at the foot 
of the area and through the tunnel, has been made of sufficient 
capacity to collect and convey the whole of the water that may 
fall when the entire collecting area of 37 acres has been laid," * 
Light, inexpensive covers are often thrown over reservoirs 
of considerable size and the sun's rays thereby excluded, with 
the result that algal growths are prevented. 
The general character of the cover that was constructed 
at Pasadena is easily determined from the illustration on page 
355, and the cost was as follows: 
259,966 square feet of lumber $4856 
9373 feet of 2-inch pipe for posts (including 551 caps 
and freight) 987 
Hardware 203 
Mill-work (labor making corbels) 27 
Cement (around foot of poles) 6 
Engineering 151 
Labor (including superintendence) 1004 
Total $7234 
The area covered was 166,000 square feet, at a cost of 4.36 
cents per square foot, or $345 per 1,000,000 gallons of storage. 
* Gibraltar Public Health Report, 1903. STORED WATER 
357 
It is important to note that in all such covers the boards 
should be laid north and south, so as to avoid too long exposure 
of the same water-surface to what sun rays may enter through 
the openings between the planks. 
With reference to the question of danger from ice formation 
the method of procedure proposed at Quincy was to " pump 
continuously into the reservoir during ice-making weather 
until full, then stop the pumps and supply the city from the 
reservoir until it was necessary to resume pumping. By this 
plan of either raising the water or allowing it to fall the ice does 
not become attached to the posts sufficiently to do any damage." 
A similar variation of water-level in a stand-pipe, during 
freezing weather might however work disaster. At a town in 
northern New York a heavy ring of ice was formed while the 
water-level was about half way up the stand-pipe. Upon filling 
the water to a high level this ring was deeply submerged. It 
subsequently became loosened from the sides, rose with great 
force and took off the cover of the stand-pipe. Cases have 
occurred where, after the formation of such an ice ring, the 
lowering of the water left it suspended in the air and its sub- 
sequent fall greatly damaged the stand-pipe sides. 
Maintaining a practically uniform water-level during cold 
weather is the reasonable procedure. 
Whatever the design of a settling-basin may be, means 
must be provided for easy removal of sediment. 
The means to be selected to clean small reservoirs will 
suggest themselves to the designer who is familiar with the 
local conditions. Gentle inclination of the reservoir floor 
toward a central sump is a usual provision. 
Before recent changes were made the Mississippi water 
used at St. Louis was improved by plain sedimentation in 
artificial basins. In all about twelve feet of mud was deposited 
therein per year, but as cleaning took place about every sixty 
days, except during freezing weather, only about two feet accu- 
mulation was permitted. The cleaning was accomplished by 358 
WATER-SUPPLY 
horse scraper aided by a siphon flow from an adjoining basin, 
except in those harder sections of the deposit which required 
hand spading as well as stream-flow. Under the conditions 
of simple sedimentation the cost of cleaning at St. Louis 
amounted to from | to 2 j cents per cubic yard of mud, according 
to its stiffness. The time for settlement varied from twenty- 
four to thirty-six hours, the lower figure being more common. 
Half of the suspended matter settled in twelve hours and 
66 per cent in from twenty-four to thirty-six hours. Not 
enough additional sediment was found to separate upon 
longer storage to make it pay to use the additional time. 
As stated in his St. Louis report (page 72) Hazen figured 
the cost of cleaning the settling basins at 15 cents per million 
gallons of water settled. This is based upon an estimate of 
fifteen cubic yards of wet mud being deposited per million 
gallons of water passing through the reservoir and a charge 
of one cent per cubic yard for its removal. This is a lower 
average figure of cost than that for 1900, for which year the cost 
was 1.49 cents per cubic yard.* 
Chemical precipitation was later added to the improvement 
process for the St. Louis water, the reagents used being sulphate 
of iron and lime. The data are as follows: f 
Average dose per gallon of water treated: 
Sulphate of iron 2.13 grains 
Lime 7.39 11 
Cost of treatment per million gallons of water treated: 
Sulphate of iron $1.442 
Lime 2.454 
3.896 
Labor 0.579 
Power 0.064 
Improvements 0.021 
Repairs 0.063 
$4,623 
* J. N. E. Water-works Asso., 19: 226. 
t Engineering Record, July 27, 1907. STORED WATER 
359 
The volume of water used for flushing in cleaning the basins 
being 2 per cent of the total water pumped. 
One form adopted for quick cleaning is pictured on page 317. 
The illustration shows the method of removing deposit from 
DEVICE FOR CLEANING A SETTLING-RESERVOIR WITHOUT THROWING IT OUT OF 
SERVICE. 
reservoirs without throwing them out of use. It will be seen 
that a series of partitions, provided with hinged covers, is placed 
in the bottom of the reservoir. The compartments thus formed 
are open at one end and closed at the other by lift-gates, which 
lead into a common header. During sedimentation the covers 
are vertical and the silt falls to the reservoir bottom. Upon 360 
WATER-SUPPLY 
closing down the shutters the deposit is trapped beneath them. 
If one or more of the gates be raised, the head of water forces 
the sediment out into the " header " and thence into the 
sewer. For larger reservoirs such a device is multiplied, and 
the shutters and gates are operated from paths of light con- 
struction. 
However imperative it is to clean sedimentation tanks or 
reservoirs at suitable intervals, no such treatment can be 
profitably considered' for those of a storage type. One might 
as properly attempt to clean the bottom of a natural lake. 
Sedimentation, of course, takes place in large reservoirs as 
well as in those of less size, but the volume of water flowing 
through them per square foot of bottom per year is but a frac- 
tion of what is stored for short intervals of time by those small 
basins, that are in comparison but enlargements of the city 
mains. 
For the removal from a large reservoir of a considerable 
growth of grass the author saw barbed wire in successful use 
at Denver, Col. One end of the wire was fastened at a point 
on shore and the other hauled by a team of horses over quite 
a sweep of bottom. The weeds were cut off near the roots or 
else dragged out bodily and afterwards floated ashore. 
" Dew-ponds " are unknown in this country but it may 
not be amiss to give the following description of them as they 
occur in England: * 
11 These ponds are familiar to those acquainted with the 
chalk downs. Many of them are of enormous antiquity, 
and are the handiwork of neolithic man. In some remote down 
districts, during long droughts, the extraordinary spectacle 
may be witnessed of villagers and farmers sending carts from the 
valley to the summit of the downs, in order to fetch water from 
dew-ponds for domestic and farm purposes. The reason that 
these dew-ponds do not dry up is that the radiation of heat from 
the earth at night is checked by a layer of non-conducting 
* A. E. Carey, in " Water, " Feb. 15, 1008. STORED WATER 
361 
medium. The surface of the pond is thus chilled to the dew 
point, and a deposition of the water vapor contained in the 
atmosphere takes place in the form of dew." 
The care of watersheds has been already touched upon 
(page 300), but a further word is here in order. 
Outside of the efforts of local authorities to safeguard the 
sources of public water-supplies, a number of State Boards 
of Health have taken up the matter and formulated rules for 
the protection of lakes and reservoirs. These rules are general 
in character and are fitted to specific cases by minor changes 
or additions to suit local circumstances. Those prepared for 
the Tomhannock reservoir supplying Troy, N. Y., read in 
part as follows: 
1. No ordinary form of privy or cesspool shall be allowed 
within 150 feet of the reservoir or entering watercourse. 
2. Privies maintained within above distance shall have 
absolutely water-tight vessels so as to, allow of removal of 
dejecta to safe location for disposal. 
3. Overflow of receptacles mentioned in (2) shall be pre- 
vented. 
4. Dejecta from the vessels mentioned in (2) shall be dis- 
posed of by burying in trenches, or digging into soil, in such 
manner as to prevent surface wash, and at a distance not less 
than 300 feet from the reservoir or 200 feet from entering 
watercourse. 
5. No sewage shall escape on ground-surface or into ground 
within 200 feet of the reservoir or of entering watercourse. 
6. No garbage, kitchen waste, laundry waste, etc., shall be 
discharged upon or into the ground within 100 feet of the 
reservoir or 75 feet of entering watercourse. 
7. No bathing shall be allowed in the reservoir or an 
entering watercourse and no animals nor poultry be allowed 
to stand, wade nor swim therein. 
8. No stable, barnyard, hog-pen, hen-house, hitching- 
place nor manure pile shall be maintained within 150 feet of 
the reservoir or 75 feet of an entering watercourse. 362 
WATER-SUPPLY 
9. Leachings from the items mentioned in (8) shall be 
deemed not properly purified unless they percolate through 
at least 150 feet of soil before reaching the reservoir or 75 feet 
before reaching an entering watercourse. 
10. No manure shall be piled nor used within 150 feet of the 
reservoir or 75 feet of entering watercourse. 
11. No fishing nor ice cutting shall be allowed on the reservoir. 
12. No human burial shall be allowed within 200 feet of 
the reservoir or 50 feet of an entering watercourse. 
Rules formulated to govern conditions on Lake Owasco 
within a distance of two miles from the intake of the city of 
Auburn, N. Y. are similar, namely: 
Least Permissible 
Distances from the 
Lake and Tribu- 
tary Watercourses. 
Feet. 
Privies, etc., of any kind that are water-tight 
Privies, etc., for permanent storage or deposit without water- 
6o 
tight receptacles 
150 
Burying human excreta, etc 
250 
House slops, sewage, bath water 
150 
Garbage, waste, matter from creamery or cheese factory, etc.. . 
80 
Stable, barnyard, hogpen, etc 
60 
Spreading compost containing human excreta 
250 
Spreading animal manure on land 
75 
Vegetable waste 
75 
Dead animals, offal, etc 
IOO 
Location of tents and other temporary shelters 
150 
When a lake is large enough to permit of navigation an 
additional rule applies, as in the instance of Skaneateles lake, 
furnishing the water-supply for Syracuse, N. Y.: 
11 No excreta, garbage, slops nor any decomposable or 
putrescible matter of any kind shall be thrown, discharged or 
allowed to escape or to pass into the lake from any steamer, 
barge, launch, sailboat or rowboat. Steamers, barges and 
other boats having water-closet or toilet accommodations 
shall be provided with removable water-tight receptacles, 
which shall be regularly emptied, cleaned and deodorized at 
least once each day." STORED WATER 
363 
These state rules are valuable so long as they remain elastic 
and retain their ability to fit local conditions by proper ad- 
justment. 
Every case must be judged upon its own merits; for the 
enactment of a too rigid general law will be likely to work harsh 
injustice on the one hand, or poor sanitary protection on the 
other. 
A " reasonable " contamination, by riparian owners, of a 
water subsequently used for public supply is permitted by 
a decision of the New York Court of Appeals, although " Nui- 
sances " are forbidden.* 
It is entirely possible to protect a reservoir, and to a great 
degree the watershed also, if careful policing of the district 
be established. Country towns situated upon the banks of 
streams tributary to the reservoir are the sources of greatest 
danger, and an earnest effort should be made to remove all 
privies, manure heaps, farmyards, refuse dumps, and other 
sources of pollution from draining into the stream or any of its 
feeders. This is sometimes difficult to accomplish completely, 
but a high degree of thoroughness can be attained by suitable 
and tactful management. 
Board of health rules are expected to cover the care of 
public watersheds, but rules and laws will not enforce them- 
selves, and the city official is commonly at a distance and more 
interested in the distribution system than in that of collection. 
The most simple arrangement would seem to be to appoint 
a local physician in each town or village upon the watershed 
as the sanitary inspector for that particular district, and to 
give him authority to employ an intelligent laborer as a sub- 
inspector to do necessary work under his direction. 
The writer had found this arrangement both efficient and 
cheap. 
The physician-inspector would be in a position to know of 
cases of disease, such as typhoid, in his district and, being so 
informed, could take proper precaution against contamination 
of the public water, which act on his part would be of vastly 
* Engineering Record, 64: 255. 364 
WATER-SUPPLY 
more practical value than a location of the trouble by some 
sanitary commission after an epidemic was well under way. 
The greatest responsibility borne by the inspector would 
be during periods when gangs of laborers were employed at 
construction work within the limits of the watershed. Should 
the workmen be many and the time of their remaining be two 
weeks or more, nothing short of incineration of all camp waste 
and night-soil should be demanded, and the utmost care should 
CONSTRUCTION OF ARMY LATRINE PIT 
Military Hygiene-Havard 
(Courtesy of Wm. Wood & Co.) 
be taken that sanitary instructions are carried out to the letter. 
A Woodruff pit, which can be constructed in a few hours, would 
be suitable for a temporary camp; while, if something more 
permanent were demanded, a well-constructed incinerating 
furnace should be built. 
The proper sanitary managing of a camp is difficult enough 
when dealing with soldiers, where military discipline aids the 
officer; but when groups of ignorant, and usually foreign, 
laborers are to be housed and fed the road to success is dis- 
tinctly harder. 
Should the camp be large and likely to be occupied several STORED WATER 
365 
months, then it would be wisdom to surround it with a man- 
proof wire fence having but a single public gateway. 
The indiscriminate use of the camp area for toilet purposes 
must be absolutely stopped, and to that end the area should 
be well cleared of all cover. 
The reservoir to be protected and its tributaries must be 
guarded against use by the camp population at all hazard and 
this may require some additional wire fencing as well as watch- 
ful patrols. 
Water for the camp should be delivered by pipe direct to 
the camp enclosure. This water should, if possible, be a gravity 
supply derived from a local source, but if it be necessary to 
use the reservoir or an entering stream, then enough water 
should be pumped to the camp to supply bathing facilities as 
well as for other uses, so as to diminish the chance of the men 
bathing in the source of supply. 
Reference has already been made to the advantage of 
destroying all human dejecta by fire, but, if such disposal should 
be impossible then the latrines should be located on as level 
ground as possible, to reduce the risk of their contents being 
scoured out by local wash. For the same reason the entire 
camp should be placed on level ground, if that arrangement 
can be made. Latrine pits, if used, should not be too deep and 
their contents should be daily sprinkled with lime and fresh 
earth and regularly protected from the reach of flies. No 
latrine pit should be allowed to fill with dejecta, but when 
the contents reach within a foot of the ground-surface the pit 
should be filled with earth and a new one dug. 
For the use of men while on the work away from camp 
a portable latrine with very shallow pit should be used and it 
should be moved frequently. Care about this point is very 
important, for the writer found that on one occasion the very 
trench in which they were laying the water-main had been used 
by the laborers for toilet purposes. 
In order to secure early information of danger threatening 
the catchment-area the Water Commissioners of one New 
England city " have adopted the plan of offering a reward to 366 
WATER-SUPPLY 
any one who reports a case of typhoid or other intestinal dis- 
ease occurring on the watershed." At Barnesville, O., pro- 
vision was made for the appointment of a local inspector, 
who must inspect the drainage area every two weeks and note 
any violation of the rules. Any case of sickness found by the 
inspector must be reported to the local board of public affairs, 
" which board shall be responsible for the care of the patient 
with respect to his causing any pollution of the water of the 
reservoir." * 
One large city goes even further: The writer is advised that 
the local physician in each town on the watershed is encouraged 
to report typhoid fever by receiving from the Water Depart- 
ment a cash payment for each case reported. Such payment 
is sufficient to cover the probable amount of his fees should 
he have had charge of the case. The patient is then removed 
beyond the limits of the watershed and cared for at the expense 
of the water department. The plan works well and avoids the 
danger of cases being concealed. 
Besides caring for the sources of pollution noted above, 
the sub-inspector's duty should include the patrol of any 
railroad track which may run near the reservoir or entering 
stream. This question of danger from railroad pollution has 
but recently been recognized. Although the possibility of 
its occurrence is always well worthy of consideration, there 
is no question but that it greatly varies in importance with 
change in topography, soil, or season. Rocky, steep slopes are 
easily washed by the rain, and frozen embankments naturally 
fall into the same classification, while flat road-beds and sandy 
soils offer better chances for polluting material to be disposed 
of by natural methods. 
Whatever the character of the road-bed, it should be the 
sub-inspector's care to remove all night-soil dropped from passing 
trains within the zone of danger, and he should exercise greater 
vigilance in winter than during summer. 
* Ohio Sanitary Bulletin, July, 1904. STORED WATER 
367 
Not every one of the people traveling by rail is in good 
health, and an appreciable number of them may be properly 
assumed to be in a condition to excrete typhoid bacilli in their 
dejecta. Should such material be dropped from the train 
at the moment when it was passing over some critical por- 
tion of a city's water system the result might be distinctly 
serious. 
It is not profitable to pay too high a price for safety, and 
the railroad danger is small, except under unusual circumstances; 
but certain precautions are warranted because reasonable both 
in method and cost. 
To the writer it seems that the remedy should be applied 
to the road-bed rather than to the train. Locking closet doors 
while passing points of danger is but a poor makeshift and fails 
under pressure of necessity in the very cases whence most 
trouble is likely to arise. As to the conversion of the car 
toilets into dry-earth closets, or equipping them with closed 
storage tanks, it is highly probable that the public would not 
approve of any such arrangement. 
It should not take unusual skill and labor to design the 
side-ditches of the railroad at dangerous points so as to catch 
what drops from the train, and cause such of it as may not 
be removed by the sanitary patrol to percolate through the 
embankment. At points of special danger, as where the line 
crosses streams near the intake, the roadway should be carried 
upon tight-decked structures, and polluting material dropped 
thereon should be carried away by suitable drains. Even dis- 
infection could be employed in cases of special risk. 
Pollution danger from railroad construction is naturally 
greater than that caused by operation, and being of the same 
type as that already considered under labor camps it should 
be cared for in the same manner. 
Highway pollution is commonly not serious, whether carried 
by wind or wash. It, of course, increases the total count of 
bacteria, but there is small danger of its containing any human 
excrement. 
Hospital drainage has seldom threatened a watershed, but 368 
WATER-SUPPLY 
the character of such an institution would instantly suggest 
that special care should be exercised in looking after the dis- 
posal of the sewage. 
The International Joint Commission, appointed to investi- 
gate the pollution of the Great Lakes and boundary waters, 
submitted a list of questions to six sanitary engineers, three 
representing the United States and three for Canada. The 
result of their work here follows: 
i. Speaking generally, water supplies taken from streams 
and lakes which receive the drainage of agricultural and graz- 
ing lands, rural communities, and unsewered towns are un- 
safe for use without purification, but are safe for use if 
purified. 
2. Water supplies taken from streams and lakes into which 
the sewage of cities and towns is directly discharged are safe 
for use after purification, provided that the load upon the 
purifying mechanism is not too great and that a sufficient 
factor of safety is maintained, and, further, provided that the 
plant is properly operated. 
3. As, in general, the boundary waters in their natural 
state are relatively clear and contain but little organic matter, 
the best index of pollution now available for the purpose of 
ascertaining whether a water-purification plant is overloaded, 
is the number of B. coli per 100 c.c. of water expressed 
as an annual average and determined from a considerable 
number of confirmatory utests regularly made throughout the 
year. 
4. While present information does not permit a definite 
limit of safe loading of a water-purification plant to be estab- 
lished, it is our judgment that this limit is exceeded if the 
annual average number of B. coli in the water delivered to 
the plant is higher than about 500 per 100 c.c., or is in 0.1 c.c. 
samples of the water B. coli is found 50 per cent of the time. 
With such a limit the number of B. coli would be less than the 
figure given during a part of the year and would be exceeded 
during some periods. STORED WATER 
369 
5. In waterways where some pollution is inevitable and 
where the ratio of the volume of water to the volume of sew- 
age is so large that no local nuisance can result, it is our judg- 
ment that the method of sewage disposal by dilution represents 
a natural resource and that the utilization of this resource 
is justifiable for economic reasons, provided that an unreason- 
able burden or responsibility is not placed upon any water- 
purification plant and that no menace to the public health is 
occasioned thereby. 
6. While realizing that in certain cases the discharge of 
crude sewage into the boundary waters may be without dan- 
ger, it is our judgment that effective sanitary administration 
requires the adoption of the general policy that no untreated 
sewage from cities or towns shall be discharged into the 
boundary waters. 
7. The nature of the sewage treatment required should 
vary according to the local conditions, each community being 
permitted to take advantage of its situation with respect to 
local conditions and its remoteness from other communities, 
with the intent that the cost of sewage treatment may be 
kept reasonably low. 
8. In general, the simplest allowable method of sewage 
treatment, such as would be suitable for small communities 
remote from other communities, should be the removal of the 
larger suspended solids by screening through a |-inch mesh 
or by sedimentation. 
9. In general no more elaborate method of sewage treat- 
ment should be required than the removal of the suspended 
solids by fine screening or by sedimentation, or both, followed 
by chemical disinfection or sterilization of the clarified sew- 
age. Except in the case of some of the smaller streams on 
the boundary, it is our judgment that such oxidizing processes 
as intermittent sand filtration and treatment by sprinkling, 
filters, contact-beds, and the like, are unnecessary, inasmuch 
as ample dilution in the lakes and large streams will provide 
sufficient oxygen for the ultimate destruction of the organic 
matter. 370 
WATER-SUPPLY 
10. Disinfection or sterilization of the sewage of a com- 
munity should be required wherever there is danger of the 
boundary waters being so polluted that the load on any water- 
purification plant becomes greater than the limit above men- 
tioned. 
11. It is our opinion that, in general, protection of public 
water supplies is more economically secured by water puri- 
fication at the intake than by sewage purification at the sewer 
outlet, but that under some conditions both water purification 
and sewage treatment may be necessary. 
12. The bacteriological tests which have been made in 
large numbers under the direction of the International Joint 
Commission indicate that in some places the pollution of the 
boundary waters is such as to be a general menace to the 
public health should the water be used without purification 
as sources of public water supply, or should they be used for 
drinking purposes by persons traveling in boats. 
13. It is our judgment that the drinking water used on 
vessels traversing boundary waters should not be taken indis- 
criminately from the waters traversed, unless subjected to ade- 
quate purification, but should be obtained preferably from 
safe sources of supply at the terminals. 
14. While recognizing that the direct discharge of fecal 
matter from boats into the boundary waters may often be 
without danger, yet in the interest of effective sanitary admin- 
istration it is our judgment that the indiscriminate discharge 
of unsterilized fecal matter from vessels into boundary waters 
should not be permitted.* 
As growing out of a consideration of reservoir-water, it is 
interesting to note the influence of the channels of distribution, 
i.e., conduits and city mains, upon the character of the supply. 
The differences between the following sets of samples, taken 
by the author at a city intake and drawn from taps at the 
other end of the town, show the changes in the water result- 
* Engineering News, July 23, 1914. STORED WATER 
371 
ing from its passing through the pumping system and street- 
mains : 
Bacteria per 
Cubic Centimetre. 
Intake. 
Tap. 
January 12 
4,022 
1502 
February 5 
3>322 
436 
April 10 
17,665 
2425 
October 30 
1,487 
lOQO 
It is instructive, and suggestive of the beneficial action of 
mains, to note that, for the year ending September 30, 1892, 
when the lake-water was pumped directly into the Chicago 
mains from the old short tunnel intakes, the percentages, by 
wards, of deaths from typhoid to total deaths were: 
Wards less than two miles from the lake 6.0 per cent 
Wards more than two miles from the lake 5.7 " 
At the time of testing the new filter for the city of Lawrence, 
Mass., the efficiency was found to be a removal of 98.3 per 
cent of all bacteria; but by the time the water reached the 
city hall 99.1 per cent had disappeared. This credit should 
be given the street-mains for the destruction of 0.8 per cent 
of the total germs. 
Similar evidence, is presented by chemical analysis of water 
from the same town: * 
Albuminoid 
Ammonia. 
Nitrogen as 
Nitrates. 
Water as pumped to reservoir 
174 
•135 
Water from reservoir. . 
T/I A 
. 146 
Water from city tap two miles from 
reservoir .117 
. 192 
In examining the Freiburg supply Tils found that the water 
from the reservoir contained fewer bacteria than that from the 
mountain source. But he also found that the bacteria increased 
in numbers after passing through the service-mains. Percy 
Frankland also found that the deep-well water furnished by the 
* Rep. Mass. Board of Health, i8go. 372 
WATER-SUPPLY 
Kent Company contained fewer bacteria as it issued from the 
wells than when delivered by the city mains to the consumers. 
This is at variance with the usual experience in this country, 
but an increase in number of ordinary bacteria during flow in 
street-mains is by no means unknown here; thus Dunwoodie 
uniformly finds such increase to take place in the mains of 
Erie, Pa. 
When securing ground for a proposed reservoir the purchase 
of all desirable building sites adjacent thereto should be care- 
fully considered, for the reason that their price is likely to be 
advanced later because of the added attraction due to the lake 
formed. CHAPTER VIII 
GROUND-WATER 
The circulation of water in the soil is governed by gravity 
and surface-tension, and the latter is in turn affected by the 
structure of the soil, its composition, and the per cent by 
volume of the empty spaces between its particles. 
The " voids " in the subsoils of South Carolina and Mary- 
land, as determined by Whitney, show as a mean for twenty- 
three localities 48.73 per cent by volume, the extremes being 
37.29 and 65.12. 
The rate at which water will flow through a soil * is depend- 
ent not only upon the pressure gradient and the aggregate 
volume of the voids, but also, and more particularly, upon their 
separate dimensions; for it can be readily seen that the inhibit- 
ing influence of friction will rapidly increase with the fineness of 
the soil-grain. 
This is observed in the following table, extracted from 
" Physical Properties of Soils." t The rates of flow through 
a certain depth are calculated for a uniform water-content of 
12 per cent. 
Soil (Maryland). 
Number of Grains 
per Gramme of 
Subsoil. 
Voids, 
Per Cent. 
Relative Time, 
Minutes. 
Pine-barrens 
. . 1.602,008x03 
40 
45 
50 
55 
55 
55 
65 
8 
Truck 
3,342,323,489 
8,258,269,975 
10.357,871,515 
11,684,097,513 
14,735,778,341 
19,638,258,585 
Tobacco 
33 
45 
49 
Wheat 
River terrace 
T riassic 
Helderberg 
IOO 
* In this connection see Wiley's " Agric. Anal.," i. 159. 
t Bulletin 4, U. S. Depart, of Agric. 
373 374 
WATER-SUPPLY 
E. L. Rogers * reported that American and French estimates 
showed the average velocity of water in sands to be about one 
mile a year, or about an eighth of an inch a minute. 
Slichter f gives an extended table of velocities under sun- 
dry conditions, from which these few instances are extracted: 
(the gradient is 100 feet to a mile, temperature 500 F., and the 
porosity 32 per cent): 
Diameter of 
Soil Grain. 
(Millimeters.) 
Velocity Miles 
per Year. 
Diameter of 
Soil Grain. 
(Millimeters.) 
Velocity Miles 
per Year. 
0.01 
0.00026 
0.80 
1.63300 
O. IO 
0.02551 
0.90 
2.06600 
O. 20 
0.10210 
1.00 
2.55IOO 
0.30 
0.22960 
2.00 
10.21000 
0.40 
0.40810 
3-00 
22.96OOO 
0.50 
0.63770 
4.00 
4O.8lOOO 
0.60 
0.91830 
5-oo 
63.77000 
0.70 
I.25000 
Professor Slichter has devised an electrical method of de- 
termining the rate of flow of ground-water which gives more 
rapid results than the chemical ones commonly employed 4 
He describes it as follows: ' (see illustration page 375). 
The up-stream well is charged with a strong electrolyte, 
which dissolves and passes down-stream with the moving water. 
The passage of the electrolyte toward the lower well is 
shown by the gradual movement of the needle of the ammeter, 
and the final arrival at the well is shown by a sudden and 
strong deflection of the needle. In an hour or two a very good 
indication of the rate can be obtained by noting the slope of the 
current curve already obtained. 
Of the numerous electrolytes tried up to the present time, 
ammonium chloride has given much the best results. With 
this salt a current of sufficient intensity to throw the needle 
of a recording ammeter can be obtained by the use of a few dry 
cells. 
* Engineering Record, 25: 4351. 
f Water Supply Paper No. 67 U. S. Geol. Sur. 
t Ibid. GROUND-WATER 
375 
The electric circuit to the wells can be made in various 
ways. A brass rod, insulated from the well tube, and lowered 
into the down-stream well by means of a rubber-covered wire, 
ARRANGEMENT OF APPARATUS FOR DETERMINING THE VELOCITY OF GROUND- 
WATER BY SLICHTER'S. ELECTRICAL METHOD. 
(A = ammeter; B = battery; R =resistance. The arrow shows direction of supposed motion 
of the ground-water.) 
may serve as one electrode, the casing of the same well consti- 
tuting the other electrode. In this case the indication of the 
movement of the ground-water will not be noted until the elec- 
trolyte has reached the lower well, where its presence will be 376 
WATER-SUPPLY 
shown by a sharp rise in the current curve. Instead of this, 
the two wells themselves may be used as the electrodes, in 
which case the gradual passage of the electrolyte down-stream 
can be observed from the beginning but the final indication 
of its arrival at the lower well is less pronounced. The best 
method is a combination of these two. The wire from the 
casing of the lower well is run to one pole of the battery with 
the ammeter in circuit. The other pole of the battery is con- 
nected both to the internal electrode of the lower well and to the 
casing of the upper well. In this case the gradual movement of 
the ground-water from one well to the other is shown on the 
ammeter chart, and an abrupt indication of the arrival of 
the electrolyte at the lower well is also recorded. With this 
device Professor Slichter has made preliminary tests of the 
rate of movement of the underflow of the Arkansas River in 
Western Kansas, and has found rates varying from 3 to 15 feet 
per day. 
The highest recorded velocity of underground water is 
said to be 144 feet in twenty-four hours, against a previous 
record of 100 feet. The new record is for water flowing through 
gravel near Tucson, Ariz. The observations were made by 
H. C. Wolff, of the Department of Mathematics, of the Uni- 
versity of Wisconsin.* 
Storer gives the following values for the water-holding 
powers of various soils. The figures show the percentage of 
water absorbed in terms of the weight of the dry soil, and 
were determined by drying, weighing, soaking, draining, and 
again weighing each sample. 
Quartz-sand, rounded edges 26 per cent 
Quartz-sand with flakes of mica... . 32 " 
Gypsum (earthy) 27 " 
Loamy clay 50 " 
White clay 74 " 
Yellow clay 68 " 
Loam 52 11 
* Engineering News, Feb. i, 1906. GROUND-WATER 
377 
Fertile marly loam 59 per cent 
Limestone-sand 29 " 
Humus 180 " 
Peat 201 11 
A word of caution seems proper here. It must be remem- 
bered that the above figures show what the sands and soils 
will hold, not what they would deliver. No pump could ex- 
tract that final portion of the contained water which would 
remain as " moisture," and its quantity would be a very re- 
spectable percentage indeed of the amounts given. 
" When a soil is only slightly moist, the water clings to its 
grains in the form of a thin film. When these soil-particles 
are brought together, the films of water surrounding them 
unite, one surface being in contact with the soil-particles and 
the other exposed to the air. If more water enter the soil, the 
film thickens, until finally, when the point of saturation is 
reached, all the space between the soil-particles becomes filled 
with water, and surface-tension within the soil is thus reduced 
to zero. Gravity then alone acts on the water and with a 
maximum force. 
" In a cubic foot of ordinary soil the total surface of the 
soil-particles will be at least 50,000 square feet. It follows that 
when the soil is only slightly moist the exposed water-surface of 
the films surrounding the soil-particles approximates that of the 
particles themselves. If such a mass of slightly moist soil be 
brought in contact with a like mass saturated with water, the 
films of water at the point of contact will begin to thicken in 
the nearly dry soil at the expense of the water-content of the 
saturated mass. The water will thus be moved in any direction. 
" During evaporation the surface-tension near the surface 
of the soil is increased, and the water is thus drawn from below. 
In like manner, when rain falls on a somewhat dry soil, the 
surface-tension is diminished, and the greater surface-tension 
below pulls the moisture down, even when gravitation would 
not be sufficient for that purpose." * 
* Bulletin 4, U. S. Weather Bureau; also Wiley's " Agric. Anal.," i. 155. 378 
WATER-SUPPLY 
Wherever found, and under whatever circumstances, the 
water of the ground owes its origin to the rain or melting snow. 
Attention is called to this point because of a widespread notion 
that the wells of fresh water often existing in the immediate 
vicinity of the ocean are fed with sea-water from which the salt 
has been removed by percolation through the sand of the beach. 
All along the shore of the Italian Riviera the traveler can see 
wells dug within 150 feet of the surf. In such cases the fresh 
HELIGOLAND 
water found originates some considerable distance inland, 
and the wells intercept it on its way towards an outlet in the sea; 
in other instances its origin is due entirely to very local rains, and 
its storage in the loose sand is owing to its being specifically 
lighter than the surrounding sea-water. 
A case of this kind is met with on a little island just east of 
Heligoland. The island is practically a mound of sand, raised 
but a few feet above the level of the North Sea, and it is per- 
haps half a mile long and two hundred yards wide. It was 
formed and is maintained by ocean currents, and is covered by " SUNK-WELL " OR " SINK HOLE " 
(Water Supply Paper No. 67) 
Underground solution-channel 
(Water Supply Paper No. 67) 
379 380 
WATER-SUPPLY 
a most scanty growth of grass. Upon this island several 
fresh water wells are maintained. Necessarily the water to be 
secured from such a source is limited in quantity. Under the 
general caption of deep-seated water we shall see that fresh 
water may reach the ocean from very distant sources, and 
under a head so great as to cause a veritable " boiling spring " 
miles out at sea. 
A commonly received conception of the occurrence of 
ground-water is that it moves in very definitely localized 
streams or veins, and that, to be successful, a well must be 
sunk so as to intercept one of_these. Of course the conforma- 
tion of the country may, at times, cause this popular notion to 
closely coincide with the truth, but a more general description 
of the occurrence of ground-water would be that of a widely 
extended sheet, and the expression water-table has been adopted 
with that view in mind. 
Underground streams, some of large size, do certainly exist, 
expecially in limestone districts, but their character would 
hardly permit of their being classed as typical " ground- 
water." If considered at all, they should be properly placed 
under " deep-seated water," although their importance as 
means of supply is entirely insignificant. 
The mean height of the " water-table " (i.e., its distance 
from the surface of the ground) is governed by the average 
rainfall and the opportunities for local drainage. The delivery 
being into the rivers and streams of the district, or into the 
sea, there is always a slight inclination of the water-surface 
toward those natural drains, more especially in their immediate 
vicinity. The seaward slope of the water-table of the south 
half of Long Island, for instance, is from 8 to 12 feet per mile. 
When a well is sunk into this layer of ground-water, and 
draught by pumping made thereupon, a cone of influence is 
established, whose apex is at the bottom of the well, and whose 
lateral elements coincide with a new and steeper slope of the 
surface of the water-table. The steepness of this slope, and 
consequently the area of the case of the 11 cone," will in large GROUND-WATER 
381 
part depend upon the character of the soil through which the 
water is caused to flow. If the grain of the soil be fine, the 
high degree of friction will greatly impede the passage of the 
water, and as a result the slope will be steep and the base of 
SECTIONS SHOWING RELATION OF GROUND SURFACE, WATER TABLE AND BED ROCK. 
(After C. S. Slichter, Water Supply Paper No. 67.) 
the cone contracted, while the reverse conditions will obtain 
in a soil of open, sandy texture. 
Throughout the semi-arid region of the great Western plains 
the ground-water and deep-seated water development have 
received a very large share of attention. If it were true that 
the " underflow," which unquestionably exists there, con- 
stantly received inexhaustible reinforcements from the moun- 
tains farther west, it would be very apparent that sterile wastes 382 
WATER-SUPPLY 
might quickly be transformed into fertile meadows by the 
sinking of wells and irrigation on an extensive scale. 
It is erroneously held by many otherwise well-informed 
people that the ground-water supplying the wells of large 
portions of the great plains of Colorado, Kansas, Nebraska, 
Wyoming, and Texas is derived from the melting of the snow 
on the Rocky Mountains; but, as is shown in the reports of 
Professors Hay and Hill,* 11 the great body of the area of the 
plains is cut off from contact with the mountains by deep river- 
trenches, which make it impossible for them to receive any 
benefit from the melting of the mountain snows." This is 
shown graphically on the next page. 
Referring to the underflow of the semi-arid region, Follett says: 
111 was detailed to collect facts bearing on the possibilities 
of general irrigation from the underground waters. Many 
facts were gathered, all tending to confirmation of the assump- 
tion that the sheet-water, broadly speaking, receives none of 
its supply directly from the mountains. This is important, as 
tending to assist in computing the possible supply. Its source 
must be the western portion of the great plains, with very little 
if any, foot-hills drainage. Here the rainfall is light, and the 
soil in general not such as will freely imbibe water, and the 
evaporation is rapid. All these conditions tend to show that 
the supply of underground water must be limited." 
He concludes that- 
" (1) The underflow is not supplied from the snow or rain- 
fall of the mountains. 
" (2) Its rate of movement in the sand is very slow, hence 
" (3) The amount which may be drawn from it without 
permanently lowering its level is small. 
" (4) Each farmer on the great plains whose land is under- 
laid by this sheet-water at a moderate depth can hope to 
obtain by pumping water enough to irrigate a small garden 
and truck-patch, say two to five acres, but 
11 (5) The supply is not such as to warrant large expendi- 
tures in constructing plants intended to obtain water sufficient 
* Senate Doc. 41, part 3, 52d Congress. GROUND-WATER 
383 
for general irrigation. Even if momentarily successful, as a 
plant would be drawing down the 
surface of a lake with no outlet, the 
supply will be exhausted. In other 
words, the water-surface will be per- 
manently lowered, and disaster to the 
irrigation-plant will follow. 
These conclusions are reached 
not only from a consideration of 
the facts here stated, but also by 
weighing many other known condi- 
tions." 
Slichter * describes " the under- 
flow " in these words: 11 The ground- 
water after starting on its journey 
toward the river valley may not after 
all find its way immediately into the 
channel. Sometimes it takes a gen- 
eral course down the thalweg and 
toward the sea within the porous 
medium itself. This movement may 
be so great as to constitute a large 
underground stream, scores of feet in 
depth and miles in breadth." Such 
a state of affairs might exist in con- 
nection with many streams, so the 
expression " underflow " must not be 
considered as applying to the western 
United States alone. The velocity 
of flow of this underground water is 
very small and is largely determined 
by the degree of fineness of the 
material through which it passes. 
The flow may, or may not, add to, 
or subtract from the water of the 
surface stream. 
Profile From the Rocky Mis West of Denver through the northern tier of Kansas 
Counties to Belleville f.of 98dMeridian. 
S' a Colorado fertiaries of the Denver Artesian Basin, b. Plains marl. c. Fertiary Grit 
* Water Supply Paper No. 67, 384 
WATER-SUPPLY 
SHOWING DELTAS OF DISAPPEARING STREAMS AS THEY LEAVE THEIR MOUNTAIN 
CANYONS 
(Courtesy of C. S. Slichter.) GROUND-WATER 
385 
Salbach * reported the water of the underflow of the Elbe 
to be distinctly different from that of the river itself; and 
that in bore holes put down through the river bottom the water 
rose six inches above the river water-level. 
When the entire surface stream is drafted off to the under- 
flow, then we have an instance of a disappearing stream such as 
may be seen in the Coast Range, California. (See page 383.) 
" Sunk wells " are at times formed by the caving in of the 
surface of the ground, and the consequent exposing of pools of 
water in a country apparently destitute of moisture. Such 
cavings are due to removal of soluble material from beneath the 
crust by the solvent action of the " underflow." Pools of this 
description have been formed in western Kansas.f (See page 
379-) 
Springs such as are so often seen issuing from the hillside 
are the primitive means by which the water of the ground 
first became available for use by man, but, inasmuch as they 
became insufficient to meet the demand, artificial openings of 
one kind or another were made whereby the buried water 
could be reached. 
When the conditions prevailing in the district do not favor 
the development of a spring on the side or at the base of a 
slope, the time-honored manner of tapping the underground 
supply is to sink the ordinary domestic well into the water-table. 
To avoid surface contamination entering the well the con- 
struction illustrated herewith J is serviceable. The well top 
is closed and is raised above the ground surface, while the joints 
of the tile-pipe casing are laid in cement down to the water- 
level. This arrangement insures a maximum filtering action 
of the intervening soil. 
The self-sinking wells at Lagos, West Africa, are constructed 
by piling rings of concrete, about one foot thick, one above an- 
* Am. Soc. C. E., xxx. 293. 
t Senate Doc. 41, part 4, 52c! Congress, p. 30. 
J See J. Sanitary Institute, xxii. 515. 386 
WATER-SUPPLY 
other, with the bottom ring arranged with a cutting edge. 
The inner diameter is about four feet and upon excavating 
the sandy soil from the interior the weight of the superimposed 
concrete rings forces the cutting edge to shear vertically while 
tne miter throws the sand inwards and the whole structure sinks 
like a caisson. No disturbance of surrounding soil thus occurs 
and consequently there is no interference with its efficiency as 
a filter.* 
TILE-LINED DOMESTIC WELL 
* J. Roy. San. Inst., xxv. 1064. 387 
GROUND-WATER 
The character of ground-water from a locality beyond the reach 
of human contamination may be judged from the following 
analysis of a spring-water from the Adirondack Mountains: 
Odor, taste and color none 
Free ammonia none 
Albuminoid ammonia 02 per million 
Nitrogen as nitrates 28 
Nitrogen as nitrites trace 
" Required oxygen " 25 
Chlorine 2. 
Turbidity none 
Alkalinity 47.5 
B. coli communis none 
Bacteria per cubic centimetre. . 1. 
Total solids 83. 
Wells of the domestic type but of large size are not in- 
frequently used for city supply; thus Schenectady, N. Y., 
secures its water from three wells each 43 feet deep. Two 
of these are circular in section with diameters of 50 feet, and the 
third, from which the water is pumped, is of " gallery " shape, 
being 60 feet long and 8 feet broad. 
The well nearest the "gallery" is located 51 feet away 
and is separated by but 7 feet from the third well. No pipe 
connection exists between these wells, the water passing 
between the three of them by percolation only. 
Upon the occasion of a large fire, when the pumps were 
run at the rate of eighteen million gallons per day, the water- 
level in the " gallery " fell 16 inches, and in the circular wells 
10 inches, and 6 inches, respectively. It is a question if the 
expense of sinking the circular wells was warranted in view of 
this experience. In all probability the " gallery " would 
receive as much, or nearly as much, inflow from the surrounding 
water-table if the wells did not exist. 
In another city the writer found a large circular well of the 
above type receiving drainage material, as indicated by the 
chemical analysis, and yet showing no pollution from a bac- 388 
WATER-SUPPLY 
METHOD OF RAISING GROUND-WATER FROM A SHALLOW WELL, TANGIER, MOROCCO. GROUND-WATER 
389 
teriological examination. The " sanitary survey " was unsatis- 
factory. This is but another instance of the risk attending the 
founding of an opinion as to the quality of the water upon one 
line of inquiry alone. 
Large wells of this kind, having great storage capacity, 
are often heavily pumped during a few hours per day and are 
then allowed to rest for refilling. By this method of use the 
" cone of influence " in the water-table assumes a steeper grade, 
the velocity of inflow increases and if there be a source of pol- 
lution in the vicinity there will be greater chance of its affect- 
ing the well-water than if the filtering power of the soil were 
not intermittently overworked. 
It is not especially rare to find waters of greatly varying 
characters at different depths in the same well flowing from dis- 
tinct, and perhaps widely separated strata. The following case 
is of a somewhat different character. A city upon one of the 
Great Lakes sank a trial well upon a sandy peninsula extend- 
ing some miles into the lake, in the hope of obtaining filtered 
lake-water therefrom. 
Samples of water from depths of 29 feet and 57 feet were 
submitted to the author for analysis, with results as follows 
(parts per million): 
Samples from 
29 Feet. 
57 Feet. 
Open Lake. 
Free ammonia 
0.28 
too high to read 
0.071 
Albuminoid ammonia 
0.189 
too high to read 
0.078 
Chlorine in chlorides 
1 ■ 5 
61. 
5. 
Nitrogen as nitrates 
trace 
trace 
none 
Nitrogen as nitrites 
0.0025 
none 
none 
Required oxygen 
5-35 
10.1 
i-75 
Total solids 
137- 
753- 
131- 
The sand appeared to be of the same quality throughout 
the depth of the well, except that the lower layers ran much 
higher in chlorides. The two waters from the well differed not 
only from each other, but also from the water of the lake itself. 
Water from the deeper portion. undoubtedly owed its char- 390 
WATER-SUPPLY 
acter to the soluble material in the sand in which it had been 
stored, while that from the more superficial sand represented 
a dilution of the same by the more or less recent rainfall which 
rested upon the denser layer below. 
Attention should be called to the fact that new wells and 
recently developed springs make but a poor showing from the 
analytical standpoint. Examinations of their waters are worth- 
less, especially from a bacteriological point of view. Gas-form- 
ing bacteria are almost certain to be discovered and altogether 
the chances for an adverse report are many. Weeks, or better, 
months, should intervene before submitting a " newly de- 
veloped " water to analysis. 
A curious instance of the contamination of ground-water 
with mineral impurity is reported by- Haworth.* In writing 
of Cherokee County, Kan., he says: "The well- and spring- 
waters before the mines were opened were first class, but as 
soon as the mines were opened all was changed, and the older 
the mines the worse the water. Animals of all kinds began 
to be seriously affected." 
The mineral deposits of the section of country above referred 
to consist largely of zinc blende, and the development of the 
mining properties permits of a ready oxidation of the zinc 
sulphide to soluble salts. Zinc-bearing spring-water from 
the neighboring portion of Missouri is reported as containing as 
much as 327 parts of zinc sulphate per million parts of water. 
It is very well known that free sulphuric acid at times 
occurs in the spring-waters of localities where pyrites is exposed 
to oxidation. 
Arsenic is occasionally a constituent of spring-water, for 
example, the water of the Maxquelle, at Durkheim, contains 
as much as 17.4 parts per million. Manganese associated with 
iron is quite common in the ground-waters of some districts, 
especially northwestern Missouri. Most instances of the 
existence of metallic salts in a water should, however, cause it 
to be classed under the general head of " mineral waters," 
and such are here manifestly out of place. 
* Am. J. Sci., xliii. 418. GROUND-WATER 
391 
For the delivery of large supplies the ground-water cannot 
by conveniently tapped by ordinary dug wells, so that recourse 
is had in such cases to what is known as 11 driven wells," set 
within suitable distance of each other, and coupled to a general 
main through which the water is drawn by the pump. 
Each well is but an iron tube, two inches to over a foot 
in diameter, perforated at its lower extremity, which is sunk 
through the soil to the water-bearing layer below. Single 
wells of this description, surmounted by a simple hand-pump, 
may be seen in some instances, replacing the domestic well in 
the country door-yard, but the type is commonly met with 
in " gangs " of considerable number of units for the supply 
of cities or towns. 
At Brooklyn, N. Y., slotted tile strainers, 20 to 40 feet 
long, are sunk in tubing previously put down, the space between 
the two filled with gravel and the casing withdrawn coincidently 
with the filling process. By the use of tile, corrosion and elec- 
trolysis are avoided and greater permanence is secured. 
A method of sinking wells by the use of live steam was 
patented a few years ago, and, under some conditions of soil, 
may show considerable saving in first cost. 
A hole some 20 feet deep is first bored with an auger, and 
in this is inserted a 6-inch heavy galvanized wrought-iron pipe, 
its lower 6 feet being perforated with |-inch holes. Inside of 
this is placed a 2-inch steam-pipe, with a nozzle formed at its 
lower end, and steam at 150 pounds pressure from a large boiler 
admitted. Sand, soil, stones, and steam escape from the 6-inch 
pipe in a continuous stream, and the pipe rapidly descends, 
being constantly turned by a man with heavy pipe-tongs 
at the mouth, and extra lengths added as necessary, until 
a supply of ground-water is found. 
By whatever method the " driven well " is sunk, its mode 
of action is entirely similar to that of the common domestic 
well, from which it differs only in diameter, and it is supplied 
by the ground-water of the district in the same manner as its 
longer-known progenitor. There is, in short, nothing gained 
in the majority of cases from the supposed exhaustion of air by 392 
WATER-SUPPLY 
DRIVEN-WELL PLANT, BROOKLYN, N. Y. GROUND-WATER 
393 
the action of the pump. Much has been claimed under this 
head, and it has been urged that the zone of influence always 
widens rapidly under " suction " from an air-tight well; but 
it must be remembered that " air-tight " is a term which can be 
usually applied to the well only, and not to the ground over- 
lying the zone of influence. The porous soil will unquestion- 
ably admit all the atmosphere required, and consequently the 
flow of water will be determined by those forces, and those 
only, which govern in the case of wells of the ordinary type. 
When, however, the well passes through an extensive layer 
of impervious clay, and taps a water-bearing stratum beneath, 
then the opportunities for a development of the advantages of 
11 suction " reach their maximum. 
If the water-level in a well be lowered by pumping, the 
capacity of the well to deliver water is progressively increased. 
For one 6-inch well under discussion, which was sunk to a depth 
of 100 feet in a 200-foot stratum of water-bearing sandstone, 
C. S. Slichter found the following delivery in cubic feet per 
minute for the several distances in feet that the water was 
lowered: * 
1 foot lowering 1.8 cubic feet delivery 
2 11 3-9 
4 " 7-2 
8 " 14.7 " 
12 " 22.1 
16 " 29.5 11 " 
20 " 36.9 
He points out the wisdom of carrying the bottom of the 
well sufficiently far below the surrounding normal water-level 
to allow of material lowering of the water in the well itself; 
and he further shows that in a well such as the above " if the 
size of the discharge pipe does not offer material resistance 
to the flow of the water " the capacity in cubic feet per minute 
will vary as follows for a varying diameter of well: 
* " Movements of Ground-Water," igth Annual Report U. S. Geol. Sur., p. 284. 394 
WATER-SUPPLY 
DRIVEN WELL (FRENCH FORM). (Garban.) GROUND-WATER 
395 
2 inch 31.90 cubic feet 
6 11 36.94 11 
12 " 41-45 " 
As cost increases with diameter these figures are worthy of 
note. Should the well be a " flowing " one under high pressure, 
then of course the item of frictional resistance in a small pipe 
casing would have to be considered. 
Great dependence is often placed upon the driven-well 
system as being an arrangement by which pure water is 
guaranteed by the thoroughness of natural filtration on a large 
scale. There is one weak point in this view which must be 
always kept in mind and guarded against. The filter is a good 
one without question, but if damaged it is beyond repair and 
it therefore should be treated with corresponding care. The 
danger is that an additional supply is frequently sought for by 
an increase in pumping capacity rather than by an extension 
of the plant. As a result the wells are over-forced, there occurs 
undue lowering of the local water-table, rapid flow of water 
toward the exhausted locality causes channel-ways to form in 
the subsoil, and surface-water consequently enters the wells 
without suitable purification. Such a condition of things being 
once established, no remedy is available. 
Writing about a large city plant, Breneman says: " Seven 
million gallons of water are daily drawn from a system of a 
hundred wells, varying in depth from 45 to 100 feet, and cov- 
ering a line about 400 feet in length. Such a yield corresponds 
to a total rainfall of 32 inches a year upon 3000 acres, or, roughly, 
represents the same annual rainfall upon all of the land within 
a radius of miles from the pumping-station. Owing to the 
sudden demand for this water the soil-waters must be con- 
tinually drawn downward in the vicinity of the pumps, and the 
nearer regions must be more effectually drained than the more 
remote. The predicted consequences are abundantly realized. 
Shallow wells in the neighborhood are wholly or nearly dry 
since the pumping-station has been opened. A swamp, for- 
merly existing about the station, has been dried up. The 396 
WATER-SUPPLY 
subsoil of a cemetery, 370 yards distant, which offers frequent 
opportunities for observation, is said by the sexton to be much 
drier than heretofore." 
A fact often lost sight of is that driven wells, so far as a 
permanent supply is concerned, are dependent upon the amounts 
of rainfall, 11 run-off," evaporation, and plant requirements. 
There is not, contrary to popular conception, an underground 
reservoir from which unlimited quantities of water may be 
pumped. It is true that a reserve storage exists, that may 
be drawn upon in time of drought, but Nature keeps a strict 
account of such matters, and the deficiency created in time of 
need must be made up during the period of plenty; other- 
wise the delivery of the plant will gradually diminish and 
ultimately entirely cease. 
The result of heavy pumping from the old Liverpool wells 
resulted in a lowering of the water-table, with infiltration 
of salt water from the river Mersey. 
Galveston, Tex., has a similar and expensive experience. 
The citizens of that city attempted to extend an excellent 
driven-well plant which they possessed by increasing the number 
of wells, and they carried the draught upon the ground-water 
beyond the point of normal supply. As a result the entire 
system was damaged by the inflow of salt water from the 
Gulf. 
One very material advantage possessed by a driven well 
over a dug one is that it can be sunk deeper in the water- 
bearing sands at small expense, and, with a long strainer, can 
take water throughout a great fraction of its length. A dug 
well, on the other hand, has its construction hampered after 
water is reached, and its cost per foot is greater beyond that 
point; so that it commonly has to depend principally upon its 
bottom for supply, tapping, as it does, only the upper portion 
of the ground-water layer. 
The form of strainer used for this type of well will vary 
from the simple " well-point " (page 394) to the more elaborate 
designs such as that described by D. H. Maury, which are 
put into place after the well-casing has been sunk to the proper GROUND-WATER 
397 
depth, the casing then being pulled back so as to uncover the 
openings of the strainer and allow the water to flow in. 
" The diameter of the holes in the inner wall is in all cases 
in., while the width of the outer 
slot is from 0.004 to 0.020 inch. 
Strainers are listed for all pipe 
diameters from 2 to 12 inches. 
" The openings in the strainer 
are made larger on the inside than 
on the outside, to prevent their 
becoming choked by sand." * 
Although channel-ways may be 
opened up in the surrounding sandy 
material by heavily pumping a 
driven-well plant, it is quite as likely 
that fine drift may pack in and 
about the strainer, reducing, if not 
practically cutting off, the delivery 
of the well, unless proper attention 
be given to cleaning it. 
The plant at Lowell, Mass., is 
cared for twice a year. Mr. R. J. 
Thomas, Superintendent, describes 
the system as consisting of 345 two and a half inch wells, 
ranging from 30 to 40 feet in depth, including the strainer at 
the bottom, which is 36 inches long. Every well has a valve 
on the pipe connecting it with the suction main and also a cap 
on its top, at the surface of the ground. 
" The first operation in cleaning a well is to close the valve 
and remove the cap just mentioned. A brush 3 inches in di- 
ameter is. then forced down by hand to the bottom of the strainer 
and drawn out again. The effect of this is to loosen the deposit 
on the sides of the well. The next step is to screw a No. 6 
Douglas pitcher pump on the top of the well and pump the same 
hard for about five minutes. Then the pump handle is raised 
and the water in the pump is allowed to run back, which together 
JOHNSON WELL STRAINER. 
* Engineering News, March 8, 1906. 398 
WATER-SUPPLY 
with water poured from pails into the well, has a still further 
cleansing effect. This is followed again by a vigorous use of 
the pitcher pump for a while. Thus the process of cleansing 
and then pumping, which we call 11 tripping," is continued until 
the water comes from the well clear. 
The total labor-cost for cleaning 345 wells is $652; cleaning 
them twice a year, as we do, of course, costs double that amount, 
or $1304. 
When preparations are being made to sink a gang of driven 
wells, consisting of a considerable number of individuals, one of 
the first questions that must be considered is the distance apart 
the wells should be placed so as not to draw from one another's 
territory. This is a point upon which no fixed rule can be 
given. In the Brooklyn plant, to which reference has been 
made, the wells are 20 feet apart, but the local conditions 
may cause this distance to be materially increased in some cases. 
It is often poor economy to place wells nearer than 50 feet from 
each other, and at times even 100 feet may be the suitable 
distance. 
A good practice to follow would be to sink two wells at 
what judgment would indicate as a proper interval, pump from 
one of these, and, if the second be too much affected by such 
pumping, increase the distance for the third well, and so on 
until the proper distance be determined. 
Closely related to the well systems already spoken of, the 
" infiltration-gallery " stands as a widely used method of 
securing the water of the ground.* 
Such a gallery is really but a dug well with one very long 
horizontal axis. Its position is usually near, and parallel to the 
banks of some stream, such a site being often chosen with a view 
of securing its supply from the water of the river. Except 
under unusual circumstances, however, the water reaching 
the gallery comes from the landward side, and is the ground- 
water of the district for which the river is the drain. 
* See pages 399 and 400. GROUND-WATER 
399 
NEWTON WATER WORKS-5 FEET BY 4 FEET 8 INCHES WOODEN COLLECTING GALLERY, 1890. 400 
WATER-SUPPLY 
Rivers may indeed diminish in volume as they flow onward, 
and may even entirely disappear by sinking into the ground, 
as we have seen, but this condition is distinctly exceptional. 
A river is commonly to be considered aS' a drain, into which 
water is received, but from which none flows. To such an 
extent is the bed of a river usually rendered impervious to 
the outward passage of water by the accumulation of fine silt, 
that an old authority goes so far as to say: " If you dug a well 
in the middle of a river, and kept out the surface-water, it is 
doubtful if you would get the river-water in your well." 
FILTER GALLERY. (Nichols.) 
The writer found the following results for water from a well 
sunk upon a sand-bank in a river, and for water from the river 
itself: 
Parts per million. 
River. 
Well. 
Free ammonia 
045 
•°45 
Albuminoid ammonia 
155 
. O 
11 Required oxygen " 
6. 
2-7 
Chlorine 
2.9 
4-3 
Nitrogen as nitrates 
337 
. 127 
Nitrogen as nitrites 
0.000 
0.000 
Total residue 
131 
ioo-5 GROUND-WATER 
401 
Also the following results for the water of a small filtering- 
gallery within 20 feet of a large river, compared with the 
water of the river itself: 
River. 
Gallery. 
Calcium carbonate 
... 26.8 
163.0 
Calcium sulphate 
... 19.4 
80.0 
Magnesium carbonate 
. . . . 0.0 
38.6 
Iron 
.... 0.7 
2-5 
Chlorine (as chlorides).... 
.... 4.0 
22 .O 
Although wells near rivers commonly do not draw river 
water, they are nevertheless liable to be contaminated thereby 
during periods of flood. 
Dependence is constantly laid upon the excellent filtering 
powers of these underground galleries, and they justify it 
during the earlier periods of their use, but, considered as a 
filter, such a device is beyond cleaning and repair; it may 
clog, or, on the other hand, ruinous channel-ways may follow 
heavy pumping. In the first instance no water, and in the 
second instance polluted water, may result. 
A development devised for the recovery of underground 
water that is chiefly found in the western parts of the United 
States is what is termed the subsurface dam. Its name is 
almost a sufficient description. If a trench extending to bed- 
rock be dug across a dry " wash " and such excavation be 
then filled with concrete, and the wall be carried upward, 
the dam so formed necessarily arrests the underflow and water 
will collect upon the up-stream side. The illustrations taken 
from 11 Water Supply Paper No. 67 " show the dam under con- 
struction and also the benefits following its completion. 
The circular collecting wells shown in the dam wall serve as 
intakes from which the distribution mains start. 
Amsterdam, like several other cities of Holland, derives 
its water from the " Sand-Dunes irregular round-topped 
sand-hills, which have been erected by wave and wind action 
along the whole of the Netherlands coast. In height they at SUB-SURFACE DAM DURING CONSTRUCTION. 
(Water Supply Paper No. 67.) 
COMPLETED DAM AND LAKE FORMED THEREBY 
402 GROUND-WATER 
403 
times reach 160 feet, but average much less, while the breadth 
of the range is very variable. Near the sea they are devoid 
of vegetation, the surface of the sand being moved to a consider- 
able degree by the action of the wind, while further inland grass 
appears and crops of various kinds are raised. 
The particular section, near Haarlem, whence Amsterdam 
draws its water, is composed of rather loose sand in which 
" reed-grass," is planted in tufts by the city water department, 
to keep the sand from shifting. The rain falling upon this 
territory each year amounts to about 28 inches. 
A natural " run-off " of the " Dunes " is very low and the 
absorption by vegetation still less; 64 per cent of the annual 
rainfall sinks into the sand. 
The area drained by the collecting system of Amsterdam 
is about 7500 acres and the volume of water secured amounts 
to some eight million gallons daily. Open canals, very similar 
to small railway cuts, wind between the sand-hills and gather 
the ground water of the district into a collecting basin, whence 
it is pumped to Amsterdam, some 14 miles away. 
The side slopes of these collecting canals are i| : 1, and 
although the cross-sections are not all of the same dimensions 
a good idea of their general size may be obtained from the 
views shown herewith. The depth of water carried in them is 
now about 6 feet. The aggregate length of the canals is about 
16 miles. 
The general level of the local water table has been lowered 
from 12 to 15 feet since 1853 by the operation of the Amster- 
dam water system; the case being similar to what has been 
observed in the chalk wells of London, namely, more water 
has been removed than could be replaced by the annual rain- 
fall. Now, however, the delivery is fixed at a figure not to ex- 
ceed the rainfall, and consequently the level of the water-table 
is held constant. 
The opportunity for the contamination of well-water, par- 
ticularly that of the common domestic well, is often great. 
On another page attention was called to the lack of a proper 404 
WATER-SUPPLY 
SAND DUNE, HOLLAND. 
COLLECTING CANAL IN SAND DUNE, HOLLAND. 405 
GROUND-WATER 
conception of the right location for the house-well among 
many of our rural people. The author has elsewhere referred to 
an instance where the well was entirely covered by a huge 
manure-heap. 
In their Sixth Report the English Rivers Pollution Commis- 
sion state the case quite graphically: 
COLLECTING CANAL IN SAND DUNE, HOLLAND. 
" The common practice in villages, and even in many 
small towns, is to dispose of the sewage and to provide for 
the water-supply of each cottage or pair of cottages upon the 
premises. In the little yard or garden attached to each tene- 
ment or pair of tenements two holes are dug in the porous soil; 
into one of these, usually the shallower of the two, all the 
filthy liquids of the house are discharged; from the other. 406 
WATER-SUPPLY 
which is sunk below the water-line of the porous stratum, the 
water for drinking and other domestic purposes is pumped. 
These two holes are not infrequently within 12 feet of each 
other, and sometimes even closer. The contents of the filth- 
hole or cesspool gradually soak away through the surrounding 
WELL SURROUNDED BY PRIVIES, RHODE ISLAND. 
soil and mingle with the water below. As the contents of the 
water-hole or well are pumped out they are immediately 
replenished from the surrounding disgusting mixture, and it is 
not, therefore, very surprising to be assured that such a well 
does not become dry even in summer. Unfortunately excre- 
mentitious liquids, especially after they have soaked through a 
few feet of porous soil, do not impair the palatability of water, 407 
GROUND-WATER 
and this polluted liquid is consumed from year to year without 
a suspicion of its character, until the cesspool and well receive 
infected sewage, and then an outbreak of epidemic disease 
compels attention to the polluted water. Indeed, our acquaint- 
ance with a very large proportion of this class of potable waters 
has been made in consequence of the occurrence of severe out- 
breaks of typhoid fever amongst the persons using them." 
The reckless manner in which a domestic well is frequently 
surrounded by sources of great pollution is here shown in an 
illustrated form from a Rhode Island case reported by E. W. 
Bowditch. 
One public well, which the author succeeded in having 
closed after much difficulty, was fouled by cesspool infiltration 
to a large extent, yet because of the coolness and sparkle of 
its water it was widely popular, so much so that its final closing 
had to be accomplished after'midnight to avoid resistance. 
As an instance of excessive pollution the following analysis 
is given of a well-water from a summer resort in northern 
New York. The water was in daily use for all domestic pur- 
purposes, was presentable to the eye, palatable and popular: 
Free ammonia .075 
Albuminoid ammonia .04 
Nitrogen in nitrates 11.4 
Nitrogen in nitrites .004 
Chlorine 40 
Total solids 566 
B. coli communis present 
Upon making the " sanitary survey," the conditions sur- 
rounding this well were found worse than even the Rhode 
Island case already quoted. 
In some cases the unsanitary arrangements causing the 
trouble receive more or less support from the law. 
In many of the towns of Ohio the local boards of health 
have determined the minimum distance to be allowed between 
a well and an uncemented privy-vault, and such distance 
has been most commonly fixed at 50 feet. The permitted 408 
WATER-SUPPLY 
distance for the town of Norwalk, a place of 8000 inhabitants, 
has been published as 25 feet, while at Bond Hill the minimum 
distance allowed is stated as 20 feet.* 
That any such distance of soil-filtration can protect a well 
from pollution, provided the polluting source be constant in 
character, is beyond even hoping for, and many instances 
could be given showing how even considerably greater dis- 
tances have also failed. 
Vaughan made sundry experiments touching upon this point; 
he says: 
11 In order to ascertain to what extent soil was contami- 
nated by privy-vaults I dug down near a privy-vault which 
was situated on the outskirts of the town and isolated, so that 
there were no other known sources of contamination around; 
I dug down a foot behind this privy-vault and took up some 
soil 3 feet below the surface to determine the amount of 
organic matter in it; then I went off 6 feet and did the same 
thing, then 12, then 18, then 24, then 30; and, without going 
into detail, suffice it to say that the contamination of the soil 
from that single privy, built upon nearly level ground, could 
be detected 50 feet from the vault plainly. This was deter- 
mined by comparing the amount of organic matter in these 
different samples of soil with other soil of the same kind where 
there were no known sources of contamination." f 
As serious a case of contaminated ground-water as can be 
quoted was responsible for the Maidstone typhoid epidemic, 
in 1897. 
Maidstone is situated upon the Medway River, but the water 
with which the town was supplied came from underground 
sources, being developments of a number of local springs so 
situated as to be utilized in three several groups. It was one 
only of these groups that had its water infected, and the cause 
of such infection is the interesting point in the case. 
The present municipal supply is derived from deep wells and 
* See Report Ohio Board of Health for 1892. 
f Ypsilanti San. Convention, July, 1885. GROUND-WATER 
409 
from galleries driven into the chalk along the plane of its 
junction with the underlying clay, and the water so secured is 
excellent, but our interest, is centered in the conditions as they 
existed in the summer of 1897. 
At that time the city's water was obtained from the ground- 
water of a highly cultivated district where the soil is often of 
stiff clay, capable of ready desiccation, with formation of deep 
soil-cracks during periods of little rain. Through such a dried 
and fissured soil water can pass rapidly, and in relatively 
large streamlets, without securing the full advantages of effi- 
cient filtration. 
1 Plenty of opportunity would be given for such polluting 
material as might be thrown upon the ground-surface to pass 
Clay puddle 
CROSS-SECTION SHOWING COLLECTING DEVICE. 
(Maidstone, England) 
the upper soil in times of rainfall following relative dryness 
and reach the tile collectors resting upon the clay bed a few 
feet below. As to what kind of pollution might occur upon 
such surface, let it be said that the whole district is heavily 
manured, and that the staple is " hops," a crop whose method 
of harvesting presents conditions both peculiar and dangerous. 
Hops ripen at Maidstone in the latter part of August, and 
the crop is gathered by a small army of " pickers " recruited 
from the slums of London. During the picking season these 
people come down in families and settle temporarily upon the 
land, a large number of them living in the open, covered by what- 
ever shelter can be secured. 
From the character of the Maidstone " pickers " it is easy 
to conceive that typhoid might be carried to the hop-fields, 410 
WATER-SUPPLY 
and, if so carried, abundant opportunity would be given for 
infection to pass into the local ground water during periods 
of drought followed by rain. Dejecta deposited upon the sur- 
face would surely be washed by the first rain into the deep 
cracks formed in the clay soil through the influence of desic- 
cation. 
In 1897 the cracks in the Maidstone soil were well developed 
SHOWING THE AMOUNTS OF DAILY RAINFALL (iN INCHES), THE VARIATIONS IN LEVEL 
OF THE SUBSOIL WATER (iN FEET) AND THE NUMBER OF TYPHOID CASES. 
The latter are charted so as to make the first of September coincide with the 
middle of August, thus allowing for the period of incubation. 
because " from June 29 to August 6 there had been no rain save 
on July 26 and 27," and then only 0.21 inch. 
Heavy rain fell on August 7 and 8, the ground-water level 
rose rapidly, and two weeks later the typhoid epidemic began. 
Dr. M. A. Adams pointed out that the sudden high-water level 
was due " to the highly desiccated clay being unable to readily 
absorb so much moisture." Later, however, " notwithstand- GROUND-WATER 
411 
ing occasional rainfall," the water level began to decline because 
of increased power of absorption on the part of the soil, and 
subsequently the heavy rains of the end of August caused 
a second rise of ground-water, which, allowing for the two 
weeks' incubation, corresponded with the climax of the 
epidemic. 
The lesson for us lies in the manifest relation existing be- 
tween a polluted surface soil and the underlying ground-water 
when the former becomes dry and fissured by prolonged 
drought. 
In view of the present interest taken in the hypochlorites 
for use in water disinfection, let it be said that subsequent 
to cutting off the infected water, the mains at Maidstone were 
flushed with a solution of " chloride of lime." The result so 
far as disinfection went was satisfactory, but the " dose " used 
was far larger than was necessary. 
No less than 583 grains of " bleaching powder " per United 
States gallon of water (equivalent to 204 grains of available 
chlorine), in other words, a 1 per cent solution, was the strength 
proposed for employment, but even this enormous dose was ex- 
ceeded through an oversight, and the solution actually used was 
still stronger. It is not surprising, therefore, to learn that 
" nearly every tap started leaking and had to be re-leathered " 
and that " the fire-engine, pump-buckets, and delivery and suc- 
tion hose were entirely destroyed." From our present knowl- 
edge of the minute dose of " bleach " that will kill intestinal 
organisms, we are perfectly willing to indorse the belief of the 
Maidstone authorities that " no microorganisms of any form 
that could have obtained access to these pipes could remain 
alive and active." 
Somewhat similar to that of Maidstone is the instance of the 
typhoid outbreak at New Harrington which resulted in 275 
cases and 26 deaths.* It appears that soil-cracks, due to 
subsidence caused by coal mines in the neighborhood, permitted 
* Report of British Med. Asso. See " Water-borne Typhoid," Hart. 412 
WATER-SUPPLY 
direct flow from farmhouse privy-vaults through some under- 
ground feeder to the town well situated some three-quarters of 
a mile away. 
Extract is here made from a report of the writer's upon 
the question of closing certain city wells: 
"As is well known, there are a number of street-pumps in 
this city, and the water which they supply is cool, sparkling, 
brilliantly clear, and generally relished. The ground into 
which these wells are sunk is the old river flood-plain, which 
extends from the present river to the eastern hills. Into this 
same soil pours the drainage from many cesspools and privies, 
CONTAMINATION OF SURFACE WELL FROM CESSPOOL. 
(Water Supply Paper No. 67.) 
and the slope of the ground in the centre of the city being 
away from the river, the natural drift of the ground-water 
toward a western outlet is to an extent impeded. In consid- 
eration of these few facts, can any reasonable person expect 
to get pure water from such a source? 
" It is a fatal error to fancy that because a water has a 
bright, sparkling, clear appearance and a pleasant taste, there- 
fore such water is wholesome. Carbonic acid gas is what 
causes the brilliancy and refreshing taste of a ground-water, 
and to the solvent action of that gas is due the clearness of 
many waters which hold much organic matter in solution. 
When it is borne in mind that carbonic acid is one of the prod- GROUND-WATER 
413 
nets of sewage decomposition, the inference as to its possible 
source in the case of the present well-waters is not a pleasant 
one. During the last four years I have at different times 
examined the waters from several of these wells, and am per- 
suaded that they are contaminated with sewage material 
beyond a peradventure. 
" It is hopeless to depend upon the purifying influence of 
the intervening soil to protect the wells from privy and cess- 
pool fouling, because soil-filtration, in order to be effective, 
must be intermittent. That is, after a ' dose ' of sewage has 
been added to a soil (and the 1 dose ' must not be a large one) 
opportunity for thorough aeration of the soil must follow, or 
the second ' dose ' cannot be purified. With a large and con- 
stant flow of polluting material the purifying powers of the soil 
quickly cease to act. 
" It will be objected that these well-waters have been in 
use for many years without bad results following. Possibly; 
but it must be remembered that the imbibition of sewage 
derived from healthy sources may be quite harmless unless it 
be in too concentrated a form, however undesirable it may be 
from an aesthetic standpoint. This has been experimentally 
proven many times. The serious part of it all is that the 
sewage which contaminates the well-water may suddenly 
become pathogenic in character and then the well becomes a 
distributing centre for disease. A city well is always to be 
suspected, and if, upon examination, its water is found impure, 
it should be forthwith ordered closed." 
No one would attempt to deny the great purifying power 
of soil; it is only the unreasonable reliance upon such power, 
and the necessity of recognizing that it is very possible to 
overload it, that are the points insisted upon here. 
Every sand-bed, and broad irrigation plant, designed for 
sewage disposal demonstrates how great is the value of soil 
filtration, but the most casual observer of those works will 
at once note that the sewage is admitted intermittently. Every 
opportunity is given for aeration and the resulting oxidation 
and nitrification. 414 
WATER-SUPPLY 
Bar wise * makes use of the following good illustration: 
Weak alcohol, allowed to trickle over twigs upon which are 
growing the Mycoderma aceti, will produce vinegar, 
c2h6o+o2=H20+C2H4O2 
VIEW OE WELL AT WALLINGFORD, CONN. 
Substituting an ammoniacal fluid, such as sewage, for the al- 
cohol, and the nitrifying germ for the above organism, a similar 
* " Purification of Sewage," 104. 415 
GROUND-WATER 
carrying of atmospheric oxygen to the oxidizable material 
will occur and nitrates will result. 
A peculiar instance of contamination of a well water occurred 
near Wallingford, Conn., A six-inch well was drilled through 
red sand-rock to the depth of sixty-six feet and water of excel- 
lent quality raised therefrom by windmill. Later an electric- 
light plant was erected, which was supplied with power from a 
gasoline engine. The necessary store of gasoline was contained 
in a cylindrical tank of riveted steel, three feet in diameter and 
ten feet long, which was buried just under the ground-surface 
at a point 130 feet distant from the well; and two gasoline 
engines were installed in a power-house nearer the well by 
thirty feet. 
About one month after the starting of the engines, a very 
decided taste and smell of gasoline developed in the water of 
the well and continued, although with diminished intensity, 
more than three years after its first being observed. 
The information furnished showed that whether from 
the exhaust of the engines, from a faulty connection or from a 
leak in the tank itself, several hundred gallons of gasoline had 
run on to or into the soil within about 100 feet of the well. 
The writer's opinion was that the injury done was beyond 
repair and that the well should be abandoned for drinking 
purposes. When one bears in mind the minute quantity of 
kerosene which remains upon the hands after handling a lamp 
and with what certainty a pitcher of ice-water is caused to 
taste of the oil if the ice be touched by the hands so soiled, 
it is easy to appreciate how very far several hundred gallons 
of gasoline would go toward contaminating a ground-water. 
Of course the day would come when the last trace of oil 
would have been washed away, but who would venture to fix 
the date for its accomplishment? 
In such a case as the above the sense of smell is more deli- 
cate than that of taste in detecting the presence of gasoline 
in water, thus it was found that 1 : 30,000 gave a strong odor 416 
WATER-SUPPLY 
but only a slight taste. Dilutions greater than that did not 
respond to taste although in a dilution of i : 180,000 gasoline 
could still be detected by the nose. 
Dilutions of kerosene as high as 1 : 1,400,000 possessed 
both slight taste and slight odor. 
An excellent way to determine the probability of objection- 
able drainage material entering a well is to place a quantity of 
a solution of common salt, of lithium chloride, or of fluorescein 
(20CH12O5) at the point whence contamination is supposed to 
come. The normal, composition of the water being known, 
there will appear an increase in u chlorides," a spectroscopic 
test for lithium, or a decided fluorescence in the water if there 
be drainage from the source in question. 
One part of fluorescein will show in 100 million parts of 
water. A convenient strength of the dye for use is 1 gramme 
dissolved in a little sodium hydroxide and diluted to one litre 
with water. Different from sundry aniline dyes which are 
sometimes used for this work, fluorescein is not decolorized 
by passing through the ground. 
Care must naturally be exercised not to alarm the takers 
of the water by the use of so intense a coloring agent, and 
where there is danger of so doing recourse must be had to the 
employment of the other indicating substances mentioned. 
When using fluorescein it must not be forgotten that some 
waters possess a slight natural fluorescence of their own.* 
It may be worth mentioning here that English and American 
laws are at variance with one another regarding the ownership 
of underground percolating water, i.e., " ground-water." 
Such ownership became definitely settled in England in 
1859 by a decision in the case of Chasemore vs. Richards, in 
which the judgment of the House of Lords was unanimous. 
" They held that the principles which regulate the right of 
* Compt. rendu, 146: 1125; also Chem. Abstracts V: 1810. GROUND-WATER 
417 
owners of land in respect to water flowing in known and defined 
channels, whether upon or below the surface of the ground, 
do not apply to underground water which percolates through 
the strata in no known channels." 
Until somewhat recently the principles laid down in this 
case have never been shaken or departed from. In every case 
the difficulty has been, not whether the decision in Chasemore 
vs. Richards was right, but rather in its application to the facts 
of the particular case. 
In a decision of the New York Court of Appeals regard- 
ing rights in underground water, in the case of Forgell vs. 
the City of New York, the decision granted damages to a 
market-gardener for the diversion of underground water from 
his land and enjoined the city from continuing to divert the 
water. 
The plaintiff owned farming land in the 26th Ward of 
Brooklyn, within about 2000 feet of the Spring Creek pumping- 
station belonging to the water-supply plant. He worked a 
market-garden and claimed that the operation of the driven 
wells at Spring Creek reduced the natural flow of water in 
ditches across his land and caused the failure of crops. Justice 
Smith of the Supreme Court decided in favor of the plaintiff, 
and awarded him $6000 damages. The city carried the case 
to the Court of Appeals, which sustained the decision of the 
lower court as to damages and also granted a perpetual injunc- 
tion restraining the city from operating its driven wells at 
Spring Creek. The court practically decided that if the city 
wanted water from driven wells, it must condemn and buy all 
the surface land affected.* 
Let us add that the law of ancient Rome allowed a man to 
dig a well in his field and thereby drain his neighbor's, unless he 
did so maliciously. 
In "Water Supply and Irrigation Paper No. 122," U. S. 
Geol. Survey, will be found a comprehensive paper prepared by 
Douglas W. Johnson which meets the demand for information 
on the relation of the law to underground waters. 
* See Engineering News, Dec. 6, 1900. 418 
WATER-SUPPLY 
It gives the common-law rules concerning underground 
waters and deals with interference, damming, artificial percola- 
tion and pollution, and concludes with legislative acts affect- 
ing such waters in a number of the States. 
Further reference to the relation of the law to water supplies 
will be found in Appendix B, page 490. CHAPTER IX 
DEEP-SEATED WATER 
Springs of small flow, such as trickle out of the country hill- 
side, are properly classified with the shallow wells already spoken 
of; they furnish " ground-water " only and are of local origin. 
Quite another matter, however, are those natural fountains 
which reach the surface in very great volume, possessed of a 
temperature radically different from that of the local subsoil, 
and holding in solution mineral materials that may be quite 
foreign to the neighborhood. Such water is of distant source, 
and the gathering-grounds where it originally falls as rain 
may be very far away indeed. 
Picture the outcrop upon some rainy upland of a porous 
stratum, encased upon either side by strata impervious to 
water; let the strata be possessed of a moderate dip, then let 
them be cut transversely at some point below, either by simple 
erosion or by a geologic fault, and the conditions for a deep- 
seated spring would be complete. Rain-water falling on the 
distant outcrop would pass down the porous stratum, picking 
up soluble material on the way, and would escape as a spring 
at the point where the strata were broken or eroded. 
Notable springs due to geologic faults occur with frequency, 
but to Americans the best-known instance is to be found at 
Saratoga, although the water furnished is medicinal rather 
than potable, and therefore beyond our present consideration. 
At the head of San Antonio River, not far from the city 
of San Antonio, Texas, is situated a mammoth spring of pure 
water, whose daily outflow is some fifty million gallons. This 
spring is but one of a group of great springs which 
11 coincide almost exactly with the line of the great Austin 
Del Rio fault." * 
* Senate Doc. 41, 52c! Congress. 
419 420 
WATER-SUPPLY 
Near Great Falls, Montana, are the " Giant Springs " 
which are estimated to flow 400 million gallons per twenty- 
four hours. The water has a temperature of 500 F. 
A curious instance of a spring of great magnitude caused 
by erosion cutting into the water-bearing stratum is to be 
found several miles out at sea off the coast of Florida, east 
of Matanzas Inlet. 
There are reasons for believing that the Matanzas spring is 
due to a bursting of the confined waters through a hole in the 
upper hard rock-layer. Successful sounding has been accom- 
plished in the spring itself. In its immediate vicinity the 
ocean suddenly deepens from a depth of 60 to one of 126 feet. 
A shipowner familiar with the locality informs the writer 
that the volume of water boiling up in the ocean at the site 
of the spring is so large as to prevent a boat remaining on it for 
more than a moment, as " the boat is washed off from it as 
from the rapids of a river." The same authority describes the 
odor of the water as that of a sulphur spring, which is an 
additional point showing its kinship to the artesian waters of 
Jacksonville and St. Augustine. 
There are several other springs on the same coast similar 
to this, although not so large. 
Fresh-water springs occur in the North Sea in the vicinity 
of the islands surrounding Holland, and are situated two or 
three miles from shore. " Similar springs may be found in 
the Adriatic Sea, near Fiume, Abazzia, Triest, and in other 
places, so that the surface of the sea is slightly raised up and 
a whirlpool may be observed." * 
Instances are not unknown where streams, or portions of 
them, disappear to rise again after subterranean flow, thus 
masquerading as true deep-seated water. 
The " source de la Loue," the largest 11 spring " but one 
in France, with a flow of 15,000 litres per second, furnishes 
an illustration of the kind; the discovery of the true character 
of the water having come about by accident in August, 1901. 
It seems that a fire occurred in an absinthe manufactory 
* Trans. Am. Soc. C. E., xxx. 300. DEEP-SEATED WATER 
421 
located upon the banks of the river Doubs, some seven miles 
distant from, and at a greater elevation than, the " spring," 
whereby great quantities of absinthe escaped into the river. 
About forty-eight hours later the water of la Loue gave 
unmistakable evidences of the presence of absinthe. 
The conclusion is that the spring represents, at least in 
part, a subterranean branch of the river Doubs. The outflow 
from la Loue again joins the Doubs lower down. 
Instances are by no means rare of the use of deep springs 
LA SOURCE DU LOIRET 
(Deep spring supplying Orleans, France.) 
for water-supplies of magnitude, such as the one which supplies 
Orleans, France, but deep-seated water is much more com- 
monly reached by special borings. 
Bore holes are made by either the diamond drill or by heavy 
chisels of solid steel. The former consists of lengths of tubing, 
ending in the cutting tool which is a short tube holding black 
diamonds staggered along the periphery of its base. Being 
rotated this tube cuts its way downward and permits of a solid 
" core " of rock being recovered for examination throughout 
the entire depth of the boring. 422 
WATER-SUPPLY 
Such drilling is commonly undertaken to determine the 
character of the strata upon or through which some engineering 
work is to be built. 
For the purpose of reaching deep-seated water, or oil, it is 
usual to employ methods using steel chisels, of great weight, 
which are raised by power and then allowed to fall. The 
rock is, of course, removed in finely divided form by this pro- 
cess. 
It would be going too far to undertake a detailed description 
of the process of drilling deep wells, yet there are certain facts 
concerning their cost and rapidity of construction, given us 
by Professor Carter,* which may properly be here inserted: 
" The most difficult rocks to drill through are trap, quartz- 
ite, compact fine-grained sandstones, certain clay slates, gran- 
ites, syenites, and compact hornblende schist, obsidian, etc. 
" The softer rocks, such as talcose and chlorite schists, 
serpentine and other magnesian rocks, limestone, dolomite, 
hydro-mica schists, and many coarse-grained sandstones, are 
readily drilled through. The following table will show the 
thickness of rock pierced by a chisel drill 20 feet long, 5! inches 
in diameter, weighing 700 pounds, guided so as to make a round 
hole: 
Locality (Pennsylvania). 
Rock. 
Rate. 
Duffield's farm, on Stony Creek, near Belfry, 
Ice company's well, Norristown 
Kunkle's farm, Valley Green Road, near 
Flourtown 
Wheadley's farm, Chester County 
Wm. Janeas' farm, near William Station... 
Roberts' well. Spring Mill 
Clay slate (Trias) 
Sandstone (Trias) 
Limestone (Silurian) 
Hydro-mica schist 
Sandstone (Potsdam) 
Sandstone (Potsdam) 
4 J ft. drilled in io hours 
5 
51 " " " 
7 
10 " " " 
i8J " 7 hours 
" The minerals which compose a rock may be very hard 
and yet the cementing material may hold the grains so loosely 
that the drill will make rapid progress through the rock. 
Sandstone, when composed entirely of silica, or when the 
cementing material is gelatinous silica, as in quartzite, is 
extremely hard to drill, but when the cement which binds the 
* J. Fk. Inst., September, 1893. DEEP-SEATED WATER 
423 
grains is feldspar, which decomposes readily, then the grains 
are loosely held, and the rock is readily drilled. 
" The price of drilling is about $2 per foot in Montgomery 
County, Pa., for wells six inches in diameter and from 100 to 
200 feet deep; this is independent of the character and hard- 
ness of the rock. 
" Other contracts in Philadelphia have been made at the 
rate of $2.75 per foot for drilling down to 500 feet, and $3 
per foot for drilling below a depth of 500 feet; this does not 
include the iron pipe for casing, but only the drilling. The 
six-inch iron pipe (internal diameter five and five-eighths 
inches) which is used to line the well varies in price from forty 
to fifty-five cents per foot." 
Wells are sunk through the shales and conglomerates of 
the Catskill Mountains, but not to great depths, at the rate of 
$2 per foot, and in the Hudson River shale of the upper valley 
at from $1.50 to $2 per foot. In the oil regions of Pennsyl- 
vania the average price is about $1 per foot for the boring alone. 
Lower prices are found in California, where the " stove- 
pipe " method of construction is in use.* 
" According to contracts recently let near Los Angeles, the 
cost of 12-inch wells was 50 cents per foot for the first 100 feet, 
and 25 cents additional per foot for each succeeding 50 feet, 
casing to be furnished free. This makes the cost of a 500- 
foot well $700 in addition to casing. The usual type of 12- 
gage, double, stove-pipe casing is about $1.05 a foot, with 
$40 for 12-foot starter with f by 8-inch steel ring. A good 
driller is paid $5 per day, helpers $2.50 per day. The cost of 
drilling will run higher than that given above in localities where 
large and numerous boulders are encountered " 
The cost of diamond drilling varies very largely with the 
diameter of the hole and the hardness of the rock. 
At Troy, N. Y., the i|-inch borings, made for inspection 
purposes along the line of the aqueduct tunnel, averaged 70 
feet deep and cost $2.25 per foot. The general character of 
the rock was hard Hudson River shale. 
* Water Supply and Irrigation Paper No. 140. 424 
WATER-SUPPLY 
The presence of boulders, as in a glacial drift, very greatly 
increases the trouble and expense of well-boring, for the reason 
that they tend to shift their position during the drilling, thereby 
throwing the hole " out of true " and causing the tools to jam. 
Logs of wood not unfrequently are encountered at great 
depths, and they always prove obstacles of considerable dif- 
ficulty. 
Bore-holes reaching to great depths: 
Feet. 
Domnitz, near Weltin, Germany 3287 
Probat-Jesar, Mecklenburg 3957 
Sperenberg, near Zossen 4173 
Unseburg, near Stassfurt 4242 
Lieth-Elmshorn, Holstein 439° 
Schladebach 5735 
Rybuik, Upper Silesia 6571 
McDonald, Penn 7181 
Galicia 7300 
" The Schladebach well was drilled under the supervision 
of the Prussian government, in search of coal. It appears that 
the total cost of drilling the well was $53,076, or at the rate 
of about $9.25 per foot of depth. The average daily rate of 
boring was 4.59 feet. The initial diameter of the hole is a 
little over 11 inches, and the least diameter in the lowest sec- 
tion is about 1.3 inches. Temperature observations which 
were made showed that at a depth of 5628 feet the tempera- 
ture was 133.8° F.* Further work on this well has been 
abandoned." 
The Rybuik well is one of the deepest in the world. A 
difficulty to contend with in so deep a boring lies in the great 
weight of the boring-tools, which in this instance reached 
30,155 pounds. Ruptures were naturally frequent and finally 
caused stoppage of the work. The cost of the boring was 
about $2.86 per foot. The temperature observations showed 
* Mechanical News, December 15, 1892. 425 
DEEP-SEATED WATER 
53l° F. at the surface and 1570 F., at the bottom, a difference 
representing an increase of about i° F. for every 63 feet.* 
Professors Chamberlin and Salisbury give the following 
records of temperature variations with depth. 
Locality. 
Depth in Feet. 
Rise of i° F. 
Sperenberg bore (Germany) 
349 2 
in 51.5 feet 
Schladeback bore (Germany) 
563° 
67.1 
Cremorne bore IN. S. Wales) 
2929 
80 
Paruschowitz bore (Upper Silesia) 
6408 
62.2 
Wheeling well (W. Va.) 
4462 
74.1 
St. Gothard tunnel (Italy-Switzerland) 
5578 
82 
M1 Cenis tunnel fFranee-Itaiv) 
5280 
79 
Tamarack mine IN. Mich.) 
445° 
IOO 
Calumet and Hecla, mine IN. Mich.) 
4939 
103 
Ditto, between 3324 feet and 
4837 
93-4 
Concerning these records the authors say: 
"It is to be noted that even these records vary a hundred per 
cent. Very notable variations are found in the same mine 
or well, and often much difference is found in adjacent records, 
especially those of artesian wells. Some of these are explain- 
able, but the full meaning of other variations is yet to be found. 
ROCK TEMPERATURES IN DEPTH ON WITWATERSRAND MINES t 
Rock Tem- 
perature at 
°F. 
1000 feet 
68.75 
2000 feet 
73-53 
3000 feet 
78.35 
4000 feet 
83.15 
5000 feet 
87.95 
6000 feet 
92-75 
7000 feet 
97-55 
8000 feet 
102.35 
General rate of increase, i° F. for 250 feet. 
The expression " artesian " has been extended to include 
deep wells under all conditions, but without proper license, 
because the term originally came from the name of the French 
* See Engineering News, Dec. 31, 1896. 
f "Geology," Vol. I., Geologic Processes and their Results, 1904. 
J South African Mining Journal, November 12, 1910, p. 417. 426 
WATER-SUPPLY 
province where deep wells were first successfully established 
in Europe, and such wells were " flowing." It is therefore to 
" flowing wells " only that the expression " artesian " properly 
CONDITIONS OF WATER-BEARING STRATA IN RELATION TO OTHER FORMATIONS. 
(After Robert Hay.) 
a, well with one water-level; b and c, wells with two water-levels; d, water-bearing gravel; 
k, impervious shale. 
CONDITION FAVORABLE FOR FLOWING WELL. (After Hay.) 
CONDITION UNFAVORABLE FOR FLOWING WELL. (After Hay.) 
attaches. Wells of this class were first sunk, in Europe, at 
Lillers, in Artois, in 1126. In the Sahara and in China they 
have been known for many centuries. Whether, however, 
the well be a flowing one or one from which the water has to 
be raised by power, the conditions, governing the storage of DEEP-SEATED WATER 
427 
such water are essentially the same as have been already given 
when speaking of deep springs. An outcrop of a porous stratum 
in a rainy upland acts as the collecting area; this stratum is of 
moderate dip and is enclosed by other strata, impervious to 
water, lying above and below. 
Unless the strata form a basin or pocket the water of the 
porous layer will find its natural outlet in spring form where 
the layer is cut transversely by erosion or fault; but should a 
well be sunk at some intermediate point the water will rise in 
the same, or overflow, to a degree dependent upon the head 
to which it is subjected and to the frictional resistance it en- 
counters; that is, to the elevation of the gathering-grounds 
above the well, and to the freedom with which the water can flow 
down the porous stratum and escape through the outlets below. 
It must be remembered in this connection that the expres- 
sion " deep " refers to the non-local character of the water 
rather than to the depth of the hole required to tap the same. 
In fact " deep water " may lie very near or even at the surface 
in some places. 
It would be difficult to find an artesian field more deserving 
of study, or more interesting to the investigator, than the one 
underlying the southeastern corner of the United States, and 
which is tapped by the wells of northern Florida, notably that 
of the Ponce de Leon at St. Augustine.* 
No one has better knowledge of that interesting well than 
Mr. W. Kennish, who was in charge during its construction. 
The following data concerning the boring are partly taken 
from private correspondence with Mr. Kennish and partly 
from his letters to the Engineering News: 
11 The pressure was found to be 17 pounds to the inch, and 
the flow 10 millions of gallons in twenty-four hours. 
11 A turbine wheel fed by this flow maintained 120 incan- 
descent lights at 16 candle-power, proving that the well was 
capable of supplying a force equal to 15 horse-power. 
* See also the excellent reports on the deep wells of New Jersey in Report of 
State Geologist for 1897 and following years. 428 
WATER-SUPPLY 
" Concerning the maintenance of the supply, we are pos- 
sessed of information upon which to form a judgment. There 
are now in the town of St. Augustine and its immediate vicin- 
ity in the neighborhood of fifty artesian wells varying in diam- 
eter from 2 to 12 inches, and exactly the same force 
exists to-day as when the first well was driven, about ten years 
since. Another ground for believing that the supply of water 
is so abundant that it will prove equal to any possible draught 
upon it by artesian wells lies in the unvarying pressure indicated 
by the very sensitive gauge of the electrical apparatus operated 
by the 12-inch well, surrounded as it is by wells on all sides 
being used in constantly varying quantities. Again, the 
increase in the diameter of the wells has been attended by 
more than a porportionate flow. 
11 While the dynamo was being operated by the 12-inch 
well a 6-inch well in its vicinity was turned on and off sud- 
denly to test the steadiness of the force, but the closest obser- 
vation did not detect the slightest trembling of the gauge. 
" From these various sources of information we cannot 
escape from the conclusion that the water under pressure beneath 
St. Augustine and vicinity is practically boundless. 
11 The ratio of increase in temperature is very nearly i° 
in every 50 feet. This increase was maintained until 86° was 
obtained at 1400 feet, where the well was stopped. 
" Between 520 and 557 feet the rock changes from fossilif- 
erous limestone to conglomerate of chalk, green clay, and a 
very porous dark fossil stone, largely composed of moulds 
of shells, the substance of which has been washed out. 
11 This stratum, of 37 feet in thickness, is doubtless caver- 
nous, through which an enormous flow passes, possibly to find 
its partial escape in the great ocean spring, five miles sea- 
ward from Matanzas, and possibly, still further from the shore, 
and at greater depth, to swell that great ocean current-the 
Gulf Stream." 
Such successive water-bearing strata as were encountered 
in the St. Augustine well are frequently observed. Thus in 
a boring at Fort Worth, Texas, at a depth of 900 feet a stream DEEP-SEATED WATER 
429 
of water was struck flowing 170 gallons per minute, with a 
pressure of 15 pounds per square inch. At a depth of 1035 
feet another lower stream was reached of 200 gallons per minute 
and pressure of 21 pounds per square inch. This second stream 
having been also cased off, further boring developed a flow at 
1127 feet in depth of 245 gallons per minute, with a pressure of 
29 pounds per square inch.* 
One interesting feature of the deep wells of northern Florida 
and of the sea-springs off the coast is the great distances the 
waters must flow in their subterranean passage from the up- 
lands of the interior, for there is no land within a long distance 
of St. Augustine of sufficient elevation to produce the " head " 
observed at the wells. 
Nearly akin to the Florida wells are those formerly furnish- 
ing the public supply for Charleston, S. C., but the pressure and 
delivery from these latter are not so great. 
The direct connection with the sea of the water-bearing 
layers which these. wells tap is shown by the water in them 
rising and falling with the tide, and it is of special interest to 
note that the daily fluctuations in the wells do not coincide 
in point of time with the tides at Charleston, a circumstance 
to be explained by the hypothesis of a distant connection 
with the sea.f 
A single case, illustrating the danger that may possibly 
follow the tapping of water-bearing sand under pressure, is 
mentioned in " Abstracts of Papers," Institution of Civil 
Engineers: J 
11 The town of Schneidenmuehl, in which the well referred 
to is situated, lies in the province of Posen, near the West 
Prussian boundary. In the autumn of 1892 the sinking of the 
well was commenced, and in May of the next year had reached 
a depth of 238 feet. From 49 to 52 feet of the upper layers 
* Senate Doc. 41, 52c! Congress, p. 106. 
f There was a famous temple of Melcarth at Gades, containing a mysterious 
spring which rose and fell inversely with the tide. (Bosworth Smith, " Car- 
thage and the Carthaginians.") 
J See also Engineering News, December 20, 1894. 430 
WATER-SUPPLY 
were comparatively firm; then a layer of sandy silt was struck, 
having a thickness of 180 feet. A current of water was used 
during the boring operations as long as the upward pressure of 
the ground-water permitted. The tubes had a diameter of 
4.72 inches to a depth of 82 feet, when they were reduced to 
3.15 inches. 
" On reaching a depth of 210 feet a strong current of 
muddy water issued from the tube. Sinking was continued, 
when at a depth of 236 feet the flow from the tube ceased; 
but a strong jet of water, amounting to 220 gallons per minute 
and containing 5 to 6 per cent of solid matters in suspension 
issued through the ground at the side of the tube. The cause 
of the hydrostatic pressure which ejected the water with such 
force is sought for in the geological formation underlying the 
town of Schneidenmuehl. This town lies in the valley of the 
river Kuedow, an effluent of the Netze, and is partially sur- 
rounded by hills composed of drift sand, which offer little ob- 
stacle to the percolation of any rain-water falling upon them. 
Upon this sand, however, lies an impervious deposit composed 
of the more or less clayey sediment left by the river, and this 
deposit appears to have resisted the pressure of the water until 
it was pierced during the sinking of the well. 
11 An endeavor was made to stop the flow of water by 
drawing the tubes, but this proved unsuccessful, because all 
solid matter which fell into the well from its sides was at once 
ejected with the water. Owing to the quantity of sand thrown 
out considerable subsidences of the surface took place, and the 
adjoining buildings were seriously damaged. Bags of sand 
and clay, stones, etc., were thrown into the well; but these 
were all swallowed up without staying the flow of water. An 
attempt was then made to dredge out a well or shaft round the 
tube, but after eight days' hard work a depth of only 2 feet 
was reached, owing to the quantity of sand driven up from 
below. It was next decided to drive down several tubes to 
the depth from which the water came, and this was done; but 
several of the tubes sank and were lost, so that finally only 
one tube remained, the minimum diameter of which was 5.9 DEEP-SEATED WATER 
431 
inches. On the 15th of June the ground sank suddenly to the 
extent of 4.26 feet, carrying with it part of the adjoining 
houses. After this the flow of water became less, and none 
reached the surface except through the tube. On the 21st of 
June tubes of gradually diminishing diameter were fixed upon 
the main tube, when it was found that the water rose to a 
height of 65.6 feet above the street-level. By means of taps 
at different levels the water was gradually cut off, and on the 
2 2d of June it had altogether ceased to flow, and the danger 
to the town appeared to have been averted. 
" On the 20th of September, however, the plug which 
closed the tube was removed,. and the water again began to 
flow, sometimes containing as much as 20 per cent fine sand. 
The water forced its way upwards outside the tube, and finally 
the fube and the whole superstructure of the well sank. The 
method devised to close the well was to heap as rapidly as 
possible sufficient earth or sand upon the spot, so that the 
weight of the layer would counterbalance the upward thrust 
of the water. An area 69 feet in diameter was cleared, and the 
six tubes were carefully filled with fine sand. The earth and 
sand which had previously been brought to the spot were 
thrown in at the rate of 1.962 cubic yards per minute until a 
conical mound was formed 6.56 feet high and 69 feet in diam- 
eter at its base. This method proved successful, and the 
issue of water from the bore-hole has now ceased." 
Flowing wells at times occur which are not due to hydro- 
static pressure, but to the lifting of local water by gas-expan- 
sion. Such a case is reported by Professor Hay as occurring 
in southeastern Kansas.* 
The same author also refers to " rock-pressure " as a cause 
of flowing ^ells of great depth. He claims that such wells 
tap sections of rock which are, together with the contained 
water, under enormous compressive strain, and the partial 
release of pressure in one direction causes the water to rise in 
* Senate Doc. 41, 52c! Congress, p. 38. 432 
WATER-SUPPLY 
Depth of well, 750 feet. 
Height of stream above ground, 80 feet. 
Diameter of well, 7 inches. 
ARTESIAN WELL AT WOONSOCKET, SOUTH DAKOTA. 
Pressure per square inch, 140 pounds. 
Temperature of water, 63° F. 
Discharge per 24 hours, 11,500,000 gallons. DEEP-SEATED WATER 
433 
the tube. This theory is combated by R. L. Jack, Govern- 
ment Geologist, Queensland.* 
A phenomenon occasionally met with in deep, non-flowing 
wells is that of "breathing." Mr. J. T. Willard, of Kansas, 
has reported an instance of that kind in which close observa- 
tion was kept of the entrance or exit of air, and the corre- 
sponding barometric readings. With a low barometer the well 
air took an outward direction, and the reverse condition followed 
increased atmospheric pressure. Mr. R. T. Smith, of Winona, 
Kan., " utilized such an air-current to blow a whistle which 
could be heard all over the town, warning the inhabitants of 
a possible storm." Of course the volume of air moving in such 
cases is far greater than what would be equal to the cubic con- 
tents of the well-tube, and it comes from a storage in porous 
or cavernous strata. 
A " breathing well " near Niagara is described in " The 
Falls of Niagara " by Spencer, Canada Geol. Sur., 1905:135. 
Attention is called by various writers-notably by Messrs. 
Todd and Swezey-to the liability of these " breathing wells " 
to freeze, owing to the sudden inflow of cold winter air from 
the outside. " The pumps not infrequently froze to the depth 
of 70 or 80 feet below the surface, and in one case ice had been 
found in a pump-cylinder 100 feet down, which was about 10 
feet above the water." 
Other things being equal, the ability of a well to furnish 
an abundant supply of water will depend upon the water- 
absorbing qualities of the rock in which the well is bored, 
and upon the number and extent of the channel-ways that may 
have formed therein. F. G. Clapp f has pointed out that 
the hard, compact and close-grained types of slate commonly 
contain more water than any other formation except sand, 
gravel, sandstone and limestone. The water is held to some 
extent in cleavage cracks, but mostly in joints, which traverse 
the rock in all directions. 
Aust. Asso. Adv. Sci., 1895. 
t Water Supply Paper No. 223. 434 
WATER-SUPPLY 
He also notes that in limestone water exists in pores and 
along joint cracks and stratification planes, but mainly in 
solution channels or cavities which have been formed by the 
slow solvent action of the waters. 11 Many of the cavities, 
but not all of them, have been formed through the enlarge- 
ment of joints; therefore they center about the intersection of 
fracture systems. Small solution channels, a few inches in 
width, and frequently many hundred feet long, can be seen in 
nearly all limestone districts, but passages of cavern size are 
found only in certain limestones. The best example in this 
country is the Mammoth Cave of Kentucky. In this cave 
some of the passages are 20 to 70 feet high and 50 to 100 feet 
broad." 
In crystalline rocks joint cracks occur as in slate, but water 
is present as well in what Clapp terms 11 concentric joints 
which are approximately parallel to the surface, and contain 
small quantities of water. They have been formed by alter- 
nate expansion and contraction of the upper zone of the rock 
owing to changes of temperature. These joints are generally 
only a foot or two apart near the surface, but many feet apart 
at greater depths." 
I 
Contrary to the belief of many people, deep-seated water 
is not inexhaustible. If the porous layers containing it be 
extensive, an immediately available supply of large volume, 
which is the accumulation perhaps of ages, may be counted 
upon; but should the daily drain be larger than the natural 
reinforcement, the delivery must surely shrink in quantity, 
and finally cease. 
Van Hise * calculates that the water in the rocks to a 
depth of 10,000 metres is equal to a sheet covering all the con- 
tinental areas 69 metres, or 226 feet deep. 
M. L. Fuller f estimates that the average per cent of rock 
occupied by water in the earth's crust is 0.52. 
He adds: "The total free water held in the earth's crust 
* Data of Geochemistry, paper 22. 
f Water Supply Paper No. 160. DEEP-SEATED WATER 
435 
would be equivalent to a uniform sheet over the entire surface 
with a depth of little less than 100 feet (96 feet), which is only 
about three-sevenths of that estimated by Van Hise. It is 
recognized that locally, where the sediments are very porous 
and of considerable thickness, several times the amount of water 
estimated may exist in unconsolidated deposits, or in the strat- 
ified rocks alone. There is a general tendency, however, to 
overestimate the amount of water in the ground owing to the 
impression of great volume which a large well often conveys, 
the fact that a large area is drained being frequently over- 
looked. The average amount of water present in the earth 
is probably under, rather than over, the amount estimated. 
It is common knowledge that a great volume of water, 
some sixty-six million U. S. gallons, is drawn daily by the 
London water department from deep wells in the chalk.* 
The serious effect of extending and heavily pumping these 
deep chalk wells supplying a portion of London is thus pointed 
out in the British Medical Journal for February 28, 1891: 
" For every two gallons of water collected within the Lee 
valley London is withdrawing three from its reservoir in that 
chalk basin, and this quite apart from the amount every day 
required by the resident population of the area. The result 
of such a process can only be a steady, if gradual, exhaustion 
of water from the chalk, and a progressive lowering of its plane 
Of saturation; and, unfortunately, facts abundantly confirm 
this calculation, and prove that such a lowering of the deep- 
water level is proceeding not merely in the valleys of the Colne 
and Lee, but in the main valley of the Thames itself as well; 
and this, moreover, to a degree which has already begun to excite 
the alarm of both agriculturists and manufacturers. 
" Springs which were perennial and abundant thirty years 
ago have now run dry; the level of the water in deep wells 
has fallen more than 20 feet within less than as many years. 
Mills are being abandoned for lack of water-power, and rivers 
which once flowed regularly past ancient mansions, built to 
* London Water-supply, Shadwell, p. 48. 436 
WATER-SUPPLY 
command a view of their bank-full streams, are now lost in 
swallow-holes, or flow only scantily, or for a few weeks in 
occasional years. In 1821 the water in a well in the east of 
London stood at Trinity high-water mark-22 feet above the 
present ordnance datum; in 1851 its average height was 43 
feet below; and in 1881 it was 105 feet below ordnance datum, 
a lowering of 127 feet in sixty years, and indicating a fall in 
the plane of saturation in the chalk of more than 200 feet in the 
century. This depletion is not explicable on any theory of a 
diminished rainfall; for in this district the average of the last 
twenty years (during which the fall has taken place at an 
increasing rate) is nearly 2 inches above the average rainfall 
for the previous thirty years. The cause is to be found wholly 
and solely in the fact that water has been, and is being, in 
creasingly drawn from the chalk basin in excess of the supply." 
Similar lowering of the deep-water table is taking place 
in this country; notably in Indiana.* 
The water-supply of Copenhagen, Denmark, is entirely from 
deep sources, reached by 5-inch wells bored through the 
glacial deposits overlying the stratum of chalk to which, or 
into which, the borings extend. Upon reaching the chalk 
layer, if it be found covered with coarse material, the well 
is not pushed to a greater depth, but should the covering layer 
be of fine grain then extension of the boring is made into the 
body of the chalk itself. Very numerous trial borings have been 
sunk throughout the country surrounding Copenhagen whereby 
complete contour maps have been made showing the undu- 
lations of the surface of the chalk deposit, referred to sea- 
level, and also giving the height at which the deep-seated 
water will rise, referred to the same datum. 
The wells are from 150 to 200 feet deep and. are sunk at 
an expense of about one dollar per foot. 
As would be expected, the water is very hard and is ill 
suited to industrial uses until softened. Such improvement is, 
* See paper by Brossman on " Water Supplies of Indiana," read before the 
Indiana Sanitary and Water Supply Association, Feb. 25, 1910. DEEP-SEATED WATER 
437 
however, always a private enterprise, the city undertaking to 
remove no mineral ingredient beyond iron carbonate, which 
latter is present in considerable quantities, 4.1 parts per million, 
or 2.3 parts Fe. 
The wells are tapped off into a common suction-main, 
whence the water is raised some 60 feet to an aerating house, 
where the soluble iron is oxidized to the insoluble form. 
This aeration and oxidation is accomplished by causing 
the pumps to deliver the water to a series of open troughs of 
cast-iron, twenty-four in number, each one being 12 feet long, 
8 inches deep and 8 inches broad. These troughs are movable 
at either end so that their level may be adjusted. The water 
overflows their sides and falls in thin sheets into the collec- 
tion well, 8 feet below, whence it flows by gravity through a 
conduit for some fourteen miles to a slow-sand filter plant situ- 
ated within the city limits. 
As the water reaches the filters it is, of course, rusty in 
appearance from the suspended iron, and the removal of this 
material is the only function which the filters have to fulfil. 
The iron remaining in the filtrate corresponds to .14 part per 
million of the carbonate. 
At the time of the writer's visit it was proposed to increase 
the time between cleanings by passing the water rapidly 
through small 11 roughing " filters of gravel and small stones 
before allowing it to flow upon the filter proper. 
No difficulty with crenothrix has been experienced. The 
filtered water is pumped directly into the distribution mains. 
Upon examining the contour maps already referred to, 
it is noticed that " cavities " in the chalk have been formed 
at the location of the several well stations and it is also ob- 
served that " cavities " in the level of the water horizon have 
been formed about the points of withdrawal, although the author- 
ities state that careful measurements have shown that no lower- 
ing of the general water-level has resulted from the operation 
of the pumps. 
The 11 deep " supply for the town of Asbury Park, N. J., 
is quite typical of the low-lying waters of that coast, and 438 
WATER-SUPPLY 
is, moreover, an instance of the successful treatment of a 
water so highly ferruginous as to be unusable in the raw 
state. 
The wells are ten inches in diameter and a little over one 
thousand feet deep and furnish a water containing 9.12 parts 
or iron per million. An " air-lift " is used to raise the water, 
which reaches the surface very red and roilly from suspended 
iron. After removal of this by simple filtration through sand 
an excellent water is obtained, clear and iron-free. The air 
is forced in through a one-inch pipe, at a depth of about two 
hundred feet and under a pressure of ninety to one hundred 
pounds. No clogging of the piping has been experienced. 
It is right to say here that although in cases like that of 
Asbury Park, where oxidation of contained iron is demanded, 
the air-lift is suited to the existing conditions, yet the fact 
should always be considered that such a device for the pur- 
pose of simply raising water is of very doubtful economy, 
except where the lift is a small one.* 
Deep water must not be expected in every locality. There 
is a widespread notion that every deep boring is sure to strike 
water in goodly quantity, if carried to sufficient depth, irre- 
spective of any surface conditions whatever. The writer has 
been called to pass judgment upon the advisability of trying 
for an artesian supply for a settlement on the top of a moun- 
tain some three thousand feet high, and that, too, a mountain 
of erosion, with horizontal strata. Shallow borings, of only 
thirty to forty feet in depth, had furnished a limited quantity 
of water in the same general locality, and hence the proposi- 
tion to increase the supply by the means above stated. The 
water obtained from the shallow wells was, of course, derived 
from the very local rainfall of the mountain-top, and was no 
indication whatever of a further deep supply, although the 
parties interested were of the opinion that water could be 
induced to run up-hill, owing to some occult " artesian " con- 
ditions. 
*See Engineering News, April 22, 1897. DEEP-SEATED WATER 
439 
Water from deep sources has, commonly, characteristics of 
its own, distinguishing it from the ground-water of the neigh- 
borhood. One of the most easily recognized of these is high 
temperature. Albertus Magnus was the first to hold that low- 
lying waters are warmed by the native internal heat of the 
earth. 
Aristotle believed the high temperature of such water to 
be due to solar heat which penetrated the crust of the earth 
and accumulated in the interior as at the focus of a lens. 
Instances of the gradual increase in temperature with depth 
of boring have already been given in data on page 425. 
Another peculiarity of deep water is the small quantity of 
dissolved oxygen it usually contains. This is by no means 
due to the pressure to which it may be subjected, for increase 
of pressure favors the solution of gases, but is rather owing to 
the abundant opportunity for removal of oxygen by contact 
with oxidizable material during the long underground journey 
of the water from its point of collection. 
The water of the Grenelle well at Paris, which flows from 
a depth of 548 metres (about 1780 feet), contains no dissolved 
oxygen whatever. 
In point of composition the waters of deep wells are almost 
always highly mineralized, as would be expected in considera- 
tion of the items of long time, long distance of flow, high pres- 
sure, and elevated temperature. Sometimes the materials 
contained render the water unfit for use, even for boiler pur- 
poses. 
E. C. Carter, chief engineer of the Chicago and North- 
western R.R., states that waters on the line in South Dakota 
from very deep sources, are so alkaline as to cause foaming. 
They are soft and very hot (1300 F.); so hot that the injectors 
will not deliver into boilers readily because the vapor breaks 
the suction vacuum. Railroad tanks which receive frequent 
renewals from these waters do not lose heat fast enough to 
keep the water cool. 
Although manifestly foreign to the present writing, it may 
be interesting to insert here the analysis of water from 11 Old 440 
WATER-SUPPLY 
Faithful " Geyser, which may be taken as typical of such 
deep waters from the Yellowstone Park.* 
NH4CI trace 
LiCl 34 per million 
NaCl 639.3 
KC1 47.8 
CsCl trace 
RbCl trace 
Na2SO4 27 
KBr 5.1 
Na2B40z 21.3 
NaAsO2 2.7 
Na2CO3 208.8 
Na2SiO3 27.9 
MgCO3 2.1 " 
CaCO3 3.8 
FeCO3 trace 
MnCO3 trace 
A12O3 1.7 ■ " 
SiO2 369.1 " 
H2S 0.2 
1390.8 per million 
Mr. Kennish furnishes the following analysis of the water 
from the Ponce de Leon well already spoken, of The water 
on issuing from the well has a strong smell of sulphuretted 
hydrogen (a quality also observed in the water of the Matanzas 
sea-spring), but all odor leaves it after standing a short time. 
Suspended matter 1.6 per million 
Silica 28.0 
Alumina 1.2 
Sodium chloride 1957 -7 
Potassium chloride 47.5 
Magnesium chloride 353-4 
* Gooch and Whitfield, Bui. 47, U. S. Geol. Sur. DEEP-SEATED WATER 
441 
Calcium sulphate 470.9 per million 
Strontium sulphate 18.6 
Magnesium sulphate none 
Calcium bicarbonate 149.9 
Magnesium bicarbonate 162.1 
3190.9 per million 
Deep waters are not always highly mineralized, however. 
Note the following: 
ANALYTICAL DATA FROM TWO DEEP WELLS OF NORTH 
CAROLINA.* 
(Note the Difference in Hardness) 
Lexington. 
Morganton. 
Odor 
none 
none 
Color 
12.5 
12.5 
Turbidity 
2.6 
i-4 
Sediment 
none 
none 
Total solids 
350. 
44. 
Alkalinity 
76.50 
26. 
Temporary hardness 
76.50 
26. 
Permanent hardness 
171 • 
17. 
Total hardness 
247.50 
43. 
Incrustants 
171. 
17- 
Iron 
• 15 
trace 
Phosphates 
trace 
trace 
Chlorine 
12 . 
6. 
Nitrates 
4 • 25 
. 32 
Nitrites 
trace 
none 
Ammonia, free 
. 01 
041 
Ammonia, albuminoid 
•03 
■ 055 
Some years ago a small sample of brown water was sent 
to' this laboratory by the gas company of Mobile, Ala. It 
came from an artesian well 600 feet deep, bored near the Gulf 
coast. A very simple and partial examination was made of it 
to determine its fitness for boiler use. It was decidedly alka- 
line from the presence of sodium carbonate, contained consider- 
able salt, and in color was of a dark coffee-brown. It is to be 
* Bui. N. C. Board of Health, April, 1907. 442 
WATER-SUPPLY 
regretted that circumstances did not favor a complete analysis 
of so interesting and unusual an artesian water. However, 
the gap has been filled by Professor Reuben Haines, who sub- 
ONE OF THE GROUP OF ARTESIAN WELLS SUPPLYING THE CITY OF JACKSONVILLE, FLORIDA. 
The water rises io feet above the surface of the ground. 
sequently reported a " remarkable artesian-well water " from 
southern Alabama.* The description of the water, and the 
depth of the well (685 feet), combined with the location, led 
* J. Fk. Inst., January, 1894. DEEP-SEATED WATER 
443 
to the conviction that the water was practically the same 
as the one previously examined by the writer. 
The analysis and comments are best given in Professor 
Haines' own words: 
11 Color (observed in tube two feet long)-very dark coffee- 
brown. 
11 Odor (heated to nearly ioo° C.)-unpleasant, odor of 
damp rotten wood. 
" Taste (warm)-disagreeable, stale, brackish alkaline with 
organic flavor. 
" Transparency-almost clear, a small amount of whitish 
sediment. 
Free ammonia 6.900 per million 
Albuminoid ammonia o. 740 
Oxygen consumed (Kubel) 10.892 
Nitrogen in nitrates 0.40 
Chlorine 998.0 
Total solid residue (dried at 1200 C.) 2060.0 
" The mineral ingredients existed in the following com- 
binations: 
Potassium sulphate 2.2 per million 
Potassium chloride 44.0 11 
Sodium chloride 1611.9 " 
Sodium carbonate 293.3 
Calcium carbonate 26.8 
Magnesium carbonate 13.4 " 
Silica, iron oxide, and alumina 8.5 " 
2000.1 per million 
" The more remarkable features of this artesian-well water 
which are here to be noted are the very high color, approxi- 
mating that of water from pools in bogs and swamps, and the 
enormous amounts of ammonia and of organic matter. Deep- 
wells and artesian waters are usually colorless and particularly 444 
WATER-SUPPLY 
free from organic matter, on account of the vast amount of 
filtration to which they are generally subjected. It does not 
seem probable that so excessive an amount of ammonia as 
occurs in the Alabama well can be referred to decomposition 
of the vegetable substance alone, but that a portion of it, at 
least, must be caused by decomposition of the organic remains 
of fish and other marine animals, which have been, perhaps, 
partially preserved from decay through the antiseptic properties 
of peat. A quantity of bones and shells mingled with sand is 
stated to have been thrown up by this well. 
" The excessive amount of sodium chloride in this well- 
water may be attributed to the probable existence of saliferous 
beds somewhere in the vicinity. From the results of analysis 
it is manifest that the water has no direct connection with the 
sea, for its mineral composition is totally unlike that of sea- 
water. The pressure of the water at the mouth of the well is 
of itself sufficient evidence that the water has an altogether 
different origin." 
Water of high color and of general swampy character is 
to be found in the deep wells in the vicinity of New Orleans, 
and is also reported in Colorado.* Only one explanation is 
apparent for all these cases, and that is that the water coming 
from its distant gathering-ground is constrained, for a portion 
of its underground course, to pass through deep-lying deposits 
rich in organic remains. 
As to what kind of rocks yield hard water and what kind 
furnish soft, we have but to consider the chemical and physical 
structure of the rock in question, and then, from the solubility 
data, a very fair judgment can be arrived at. Professor Carter 
has summed up this question very aptly.f He says. 
" Water that passes through calcareous or magnesium 
rocks of great thickness will probably be hard, while water 
that passes through rocks composed of silica, alumina, iron, 
potash, or soda will probably be soft. The great deposits of 
* J. Am. Water-works Asso., 1897, p. 137. 
f J. Fk. Inst., September, 1893. , DEEP-SEATED WATER 
445 
limestone, marble, gypsum, and other calcareous rocks, as 
well as the magnesian rocks, such as dolomite, chlorite, and 
talcose schists, would then yield hard water. The granites, 
gneisses, and many sandstones and slates would furnish soft 
water. This, in the main, is true, although there are some 
exceptions which local conditions modify. Some sandstones 
will yield soft water, while other sandstones furnish hard water; 
it depends mainly upon the cement which binds the grains 
together. If the cementing material be carbonate of lime or 
sulphate of lime, the water will probably be hard, especially if 
the well be of great depth and the water be long in contact 
with the rocks. If, on the other hand, the cementing material 
be feldspar, such as orthoclase or albite (not labradorite), or 
even gelatinous silica, the water will probably be soft. 
11 Water which flows through calcareous channels is hard, 
while that which flows through silicious rocks is soft." 
As has been shown, deep water may at times be too highly 
mineralized, or even too " peaty," for potable use, but such 
impurity is not the only form that may be present on occasion. 
Even a deep well, especially if not a true flowing artesian, 
is not always exempt from local infiltrations of contaminating 
character. The writer condemned a deep-well water from Erie, 
Pa., upon the following analysis: 
Free ammonia 2.025 
Albuminoid ammonia none 
Chlorine 69 
Nitrogen as nitrates .025 
Nitrogen as nitrites none 
" Required oxygen " .85 
Total solids 487 
Phosphates strong traces 
The well had been bored within city limits, through a friable 
rock, and within 75 feet of the nearest privy-vault. 
Adverse report was also made upon another well, which 
had been very carefully bored through rock (shale) for 200 446 
WATER-SUPPLY 
feet. Within 50 feet of the well was a large privy-vault. 
Sixty-five pounds of salt, .dissolved in three barrels of water, 
were thrown into the vault, and the water of the well was 
watched for increased chlorine, with the following results: 
Before 11 salting " 58 parts chlorine per million 
After 15 hours'pumping 64-25 
"36 " 64.37 " " " 
Careful lining of the well with iron pipe will not with cer- 
tainty safeguard the water from contamination. No better 
illustration of this truth can be had than the circumstances 
attending the typhoid disaster at the school already given, where 
a broken sewer-drain permitted pollution to run down the annular 
space between the rock wall and the outer side of the pipe. 
From the bottom of the well so infected the water was pumped 
for drinking purposes. 
The mere fact that the well is deep is commonly considered 
to insure safety, and fortunately the trust is usually not mis- 
placed, but much depends upon the character of the material 
into which the well is sunk. Chalk or limestone wells are 
sometimes unreliable because of the highly-fissured formations 
likely to be encountered, and again there are found, as stated 
above, deep-well waters polluted by surface material when the 
drilling has been done in friable material, such as Hudson River 
shale. 
Waters from deep sources are not to be rated as " uniformly 
sterile." Leaving out of consideration such instances as wells 
which permit the direct entrance of surface drainage, there are 
yet many deep waters showing abundance of bacterial life; 
but, fortunately, the chances of encountering therein germs 
of objectionable character are reduced to a-minimum. Prof. 
Sedgwick has reported his findings upon this question: 
11 From our results we are forced to the conclusion that 
ground-waters, even the waters of deep wells, may not be by 
any means as free from bacteria as has been hitherto supposed. 
"It is plain that water absolutely free from bacteria is not 
ordinarily obtained from even deep wells, and that many deep DEEP-SEATED WATER 
447 
wells contain as numerous bacteria as are found in many sur- 
face-waters." 
The bacteria present in these waters are, however, " remarK- 
able not only for slow growth, but also for the absence of 
liquefying colonies, and, in many cases, for the abundance of 
chromogenic varieties. These facts are especially important as 
indicating the total absence of contamination by ordinary sur- 
face-water, and, as far as they go, they strengthen the confidence 
with which well-protected ground-waters may be regarded as 
sources of public water-supplies." * 
Kellerman and Whittaker f report finding B. Coli in 1 c.c. 
samples of water from 6 out of 13 rock-drilled wells examined. 
It should be borne in mind that if deep wells become by chance 
polluted, the probable absence of competing organisms would 
give those of pathogenic character a better opportunity to 
become dangerous. 
Before concluding this chapter on deep-seated water the 
writer begs to introduce a somewhat extraneous matter, that 
can claim attention only because of its unusual character. 
Kephalonia, the largest of the Ionian Islands, lies off the west 
coast of Greece and is some thirty miles long by twenty miles 
broad. Near its principal town of Argostoli are mills equipped 
with undershot wheels which are driven by sea-water flowing 
landward from the coast. 
In the words of Baedeker, 11 The mills are driven by a current 
of sea-water which flows into the land for about fifty yards 
finally disappearing amid clefts and fissures in the limestone 
rock." 
The illustration shown herewith is of the " Old Mill " 
erected in 1835, and shows the water flowing in the tail-race 
towards the point of its mysterious disappearance, which is 
some ten feet below the level of the sea. The question is, Where 
does the water go? 
During the British occupancy of the island, pitch, petroleum 
and other easily noticeable substances were poured into these 
* Sedgwick, Rep. Mass. Board of Health, 1894, p. 443. 
f U. S. Dept. Agri. Bui. 154. 448 
WATER-SUPPLY 
" clefts and fissures " and a strict watch of the surrounding sea 
failed to detect their reappearance. 
Is it not possible that the water which sinks into the rocks 
of Kephalonia comes to the surface again in the form of steam 
at Stromboli, Vesuvius or other volcanic vents? 
Crosby suggests another explanation, namely, that an extinct 
SEA-MILL, KEPHALONIA, GREECE. 
volcanic vent at Kephalonia receives the water and passes it 
to another similar arm of a mighty U. The water after reach- 
ing the lowest level becomes heated and then rises in the second 
arm of the U. Under such conditions constant flow in one direc- 
tion would be established and the discharge might be made at 
some distant point of the sea bottom. CHAPTER X 
QUANTITY OF PER CAPITA DAILY SUPPLY 
The following table is from a paper by Nicholas S. Hill, Jr.,* 
giving data as to 
PER CAPITA WATER CONSUMPTION IN UNMETERED CITIES OF 
THE UNITED STATES 
Daily Consump- 
tion per Capita. 
(U. S. Gallons.) 
Connecticut: 
Bridgeport 
200 
Hartford 
68 
Meriden ................... 
TOO 
New Haven 
150 
New London 
155 
Stamford 
150 
Waterbury 
91 
New York: 
Albany 
220 
Auburn 
182 
Binghamton 
140 
Brooklyn 
91 
Buffalo 
3°9 
Elmira 
122 
Ithaca 
133 
Kingston 
266 
Newburgh 
144 
New York (Manhattan) 
104 
Oswego 
196 
Poughkeepsie 
03 
Rochester 
04 
Schenectady 
123 
Syracuse 
I 20 
Troy 
271 
Yonkers 
06 
New Jersey: 
Atlantic City 
24 s 
Camden 
I 2< 
Elizabeth 
164 
Jersey City 
156 
Newark 
102 
*J. Am. Water-works Assoc., March, 1915. 
449 450 
WATER-SUPPLY 
Daily Consump- 
tion per Capita. 
(U. S. Gallons.) 
New Jersey: 
Passaic 
72 
150 
191 
251 
94 
117 
153 
160 
120 
80 
178 
221 
138 
66 
Plainfield 
Trenton 
Pennsylvania: 
Allegheny 
Altoona 
Bradford 
Erie 
Johnstown 
Lancaster 
Norristown 
Philadelphia 
Pittsburgh 
Reading 
York 
Average 
161 
PER CAPITA CONSUMPTION OF WATER IN CITIES HAVING A LARGE 
PERCENTAGE OF METERED SERVICES 
Per Cent 
Metered 
Services. 
Daily Con- 
sumption per 
Capita. 
(U. S. Gallons.) 
Anniston, Alabama 
76 
128 
Atlanta, Georgia 
IOO 
Bayonne, New Jersey 
IOO 
95 
Bessemer, Alabama 
70 
46 
Brockton, Massachusetts 
00 
28 
Cleveland, Ohio 
04 
IOO 
Consolidated Water Company 
8^ 
71 
Covington, Kentucky 
IOO 
51 
Geneva, New York 
01 
Harrisburg, Pennsylvania 
83 
122 
Hartford, Connecticut 
08 
68 
Lexington, Kentucky 
98 
58 
Lincoln, Nebraska 
IOO 
4° 
Milwaukee, Wisconsin 
99 
IO< 
Montclair, New Jersey 
IOO 
66 
Newton, New Jersey 
86 
eg 
New York Inter-Urban Water Company 
95 
78 
Passaic, New Jersey 
54 
72 
Providence, Rhode Island 
9° 
68 
Rochester and Lake Ontario Water Company 
IOO 
94 
Utica, New York 
98 
^0 
Yonkers, New York 
Average 
IOO 
96 
72 QUANTITY OF PER CAPITA DAILY SUPPLY 
PER CAPITA CONSUMPTION IN THIRTY-SIX ENGLISH 
PROVINCIAL CITIES 
(London County Council Report, 1897) 
451 
Average Daily Supply, 
U. S. Gallons, per 
Capita. 
Barrow 
42 
Birkenhead 
31 
Birmingham ...... 
28 
Bolton 
26 
Brighton 
43 
Bradford 
3i 
Burnley 
26 
Bury * . . . 
29 
Cardiff 
29 
Coventry . 
28 
Croydon 
35 
Derby 
27 
Gloucester 
21 
Halifax ., 
24 
Huddersfield 
30 
Hull 
49 
Ipswich 
21 
Leeds 
43 
Leicester -. 
22 
Lincoln 
28 
Liverpool 
34 
Manchester 
40 
Middlesborough 
61 
Newport 
27 
Northampton 
23 
Nottingham 
24 
Oxford 
3° 
Plymouth 
59 
Reading 
42 
St. Helens 
49 
Salford 
26 
Sheffield 
21 
Southampton. 
45 
Swansea 
35 
Wigan 
20 
Worcester 
43 
Average. 
00 452 
WATER-SUPPLY 
PER CAPITA CONSUMPTION IN GERMAN CITIES 
(After Brackett) 
Place. 
Population. 
Daily Con- 
sumption, 
U. S. Gals. 
Place. 
Population. 
Daily Con- 
sumption, 
U. S. Gals. 
Altona 
156,500 
118,500 
74,5oo 
1,606,424 
52,000 
335,ooo 
139,374 
255,000 
105,712 
107,085 
280,200 
i55,9oo 
137,000 
186,000 
48,200 
583,700 
26.07 
33-59 
34-78 
16.37 
24-94 
21.71 
11-50 
45-22 
18.52 
26.70 
21-54 
22.10 
29.92 
36.26 
41.46 
58.00 
Hanover 
Halle.. 
189,976 
120,000 
74,200 
72,000 
162,000 
198,000 
298,000 
145,000 
70,000 
118,000 
139,200 
66,000 
61,032 
96,650 
T8.40 
21-94 
28.14 
20. r8 
16.87 
25-24 
34-oo 
I7-4I 
13-33 
31-54 
21-34 
20.74 
35-50 
56.71 
Barmen 
Basel 
Karlsruhe.. . . 
Kiel.. 
Berlin 
Bonn 
Konigsberg . . 
Magdeburg. . 
Munich.. . 
Breslau 
Chemnitz 
Cologne 
Nuremberg . . 
Posen 
Crefeld 
Danzig 
Stettin . . . 
Dresden 
Stuttgart.... 
Wiesbaden.. . 
Wurzburg.. . . 
Zurich 
Dusseldorf 
Elberfeld 
Frankfort 
Freiburg, B.. . . 
Hamburg 
Average. . . 
27.69 
The nine large conduits of Rome at the time of Nero de- 
livered 173,000,000 gallons daily. Afterwards the increased 
supply furnished 312,000,000 gallons daily, or over 300 gallons 
per capita per day.* 
Upon glancing over such data as have been given for cities 
of the United States, and bearing in mind how often the water 
furnished our towns is inferior in character, one is impressed 
with the thought that we Americans are often more concerned 
about the quantity of the supply than about its quality. 
There is no question but that our allowance is unreason- 
ably large. Fifty gallons is considered a generous amount per 
individual in Europe, but it would be deemed quite a small 
quantity here in America. 
If we had but an increased cleanliness to show for our 
great use of water, there would be a measure of compensation 
for the additional cost, but the writer confesses to an inability 
* Senate Doc. 41, 52d Congress, part 1, p. 431. For estimates of Forbes and 
Clemens Herschel, see p. 10. QUANTITY OF PER CAPITA DAILY SUPPLY 
453 
to detect wherein our American cities are superior to those of 
Europe in this particular. 
The question of what is a suitable quantity of water to 
allow for daily consumption per capita is a very vexed one and 
great variety of opinion is on record. " Water should be as 
free as air " is insisted upon by some, while their opponents, are 
strong in the belief that an individual consumer has no more 
right to unnecessarily increase the public charge for water 
than he has to ask his neighbor to help pay for his waste of 
illuminating gas. 
" Plenty of water helps to flush the sewers " is advanced 
by others as an argument for a large per capita allowance; 
but whoever has seen the small volume of sewage that commonly 
flows along the bottom of a sewer can well understand how 
slight a chance for efficient flushing is furnished by any reason- 
able increase of per capita water allowance. 
It would seem that an allotment of 100 gallons per capita 
per day should be considered liberal and yet, in view of the 
wasteful habits that our water consumers have acquired, engi- 
neers are forced to be somewhat cautious in designing for a 
supply based upon that reasonable figure. 
In estimating the future needs of the city of Baltimore those 
in charge of the new works concluded it safer to assume that the 
following per capita daily supply would be required: 
1915 130 gallons 
1920. . 135 " 
i93° 145 11 
1940 15° 11 
As cities grow in population and as the mileage of sewers 
and the introduction of plumbing is extended the amount of 
water used and wasted per individual inhabitant also increases. 
Each use of an ordinary bath-tub requires about 20 gallons 
of water and every emptying of a toilet tank calls for 4 gallons 
in addition. Possession of sewer facilities naturally increases 
plumbing conveniences with their inevitable waste of water. 
The enlargement of the per capita supply occasioned by the 454 
WATER-SUPPLY 
introduction of public water and a system of sewers is well shown 
by the following data published by the Mass. Board of Health 
Report, 1900, page 607: 
" Table showing the increase in the consumption of water 
per person with the age of the works. (Average of twenty- 
four cities and towns.) 
Year After Water Was 
Introduced. 
Average Con- 
sumption per 
Person per Day. 
Year After Water Was 
Introduced. 
Average Con- 
cumption per 
Person per Day. 
First 
23 
Sixth 
37 
Second 
28 
Seventh 
39 
Third 
3° 
Eighth 
40 
Fourth 
32 
Ninth 
42 
Fifth 
36 
Tenth 
43 
11 This table shows an increase, in ten years, of nearly 100 
per cent. 
"Average consumption of water in three cities before and 
after the introduction of a fairly complete system of sewers." 
City. 
Years Previous to Introduc- 
tion of Sewers. 
Years Subsequent to Introduction 
of Sewers. 
7 
6 
5 
4 
3 
2 
I 
I 
2 
3 
4 
5 
6 
7 
8 
9 
Marlborough.. .. 
13 
17 
20 
21 
24 
24 
25 
30 
30 
35 
34 
37 
37 
38 
38 
36 
Newton 
28 
3i 
33 
33 
3i 
36 
4° 
50 
52 
60 
65 
63 
60 
57 
63 
62 
Waltham 
37 
36 
39 
33 
3i 
32 
33 
47 
53 
61 
59 
7i 
70 
76 
88 
9° 
Analysis of the sundry items included under the general 
head of " Water for Public Purposes * 
U. S. Gallons per Capita 
per Day. 
Public buildings, schools, and hospitals 2.30 
Street-sprinkling 1.00 
Flushing sewers and public urinals 10 
Ornamental and drinking fountains 25 
Fires 10 
Total for public purposes 3.75 
* J. N. E. Water-works Assoc., xi. 71. QUANTITY OF PER CAPITA DAILY SUPPLY 
455 
" Probably 4 or 5 gallons per capita should cover all re- 
quirements for public purposes." 
( The useless waste of water in our cities is something enor- 
mous. In Chicago, Cleveland, Philadelphia, and Detroit 
the probable waste has been estimated at about 50 per cent, 
while in Buffalo the enormous figure of 70 per cent was given 
by the city engineer. ) 
That the great bulk of this waste could be saved by meter 
measurement is an already demonstrated fact, and the fixing 
of a minimum daily allowance of water, for which the con- 
sumer would have to pay, whether he used it or not, would 
remove the objections that might be raised to meters upon 
sanitary grounds. The author has corresponded with city 
health officers in various parts of the country, with a view of 
determining what, if any, is the effect of the meter system 
upon public health, arising from an attempt on the part of the 
poorer classes to economize in the use of water. The reply 
from Providence, R. I., is quite typical: " I do not find that it 
diminishes the proper use of water in the slightest degree. Its 
only tendency is to diminish waste. There is in my opinion 
no objection, from a sanitary point of view, to the use of 
meters." 
In Providence, R. I., the minimum charge was $10 per family 
per year, which entitled them to 91.32 gallons per family per 
day, or at five persons per family to 18.26 gallons per capita 
daily. 
An analysis of the water accounts of 2553 families who paid 
the minimum rate, but who were " of a good class " showed 
the following results: 
167 families used less than 6.15 gallons per capita daily. 
237 11 11 " 8.20 
361 " " " 10.25 " " " 
445 " " " 12.30 
446 11 " " . 14 35 
462 " 11 11 16.40 " " " 
.435 " 11 11 18.27 " . " 456 
WATER-SUPPLY 
Hence the bulk of these people did not use all for which they 
were paying.* 
It is often assumed that the water required for fire service 
represents a large fraction of that charged to per capita con- 
sumption, and this may readily be true for small places; but 
as cities grow in size the draft made by the fire engines becomes 
progressively smaller per individual citizen. Placing 250 gallons 
per minute as the water needed for a fire stream (the hydrant 
pressure being say 75 pounds) and allowing ten such streams 
for a fire, the sum total of water thrown does not represent a 
serious addition of cost to each taxpayer in a large community. 
In this connection it is interesting to note that during the 
great fire at Baltimore when over fifty engines were at work, 
and when further loss of water occurred through the breaking 
of thousands of service pipes, the flow of water exceeded the 
average daily consumption by only 82 per cent.f 
Discussing the question of fire service, Mr. Nicholas S. 
Hill, Jr. says:! 
" When fire service is furnished directly from the mains, 
service pressures of not less than 70 to 75 pounds at the hydrant 
under conditions of maximum draft should be maintained in the 
congested value districts, although 50 to 60 pounds minimum 
will usually suffice for residential sections. With these pressures, 
fire streams throwing from 200 to 250 gallons per minute through 
i|-inch smooth nozzles attached to 200 feet of 2|-inch ordinary, 
best quality, rubber-lined hose may be obtained. 
" Where fire engines are used, a service pressure of 30 pounds 
minimum on the mains will usually be ample in residential sec- 
tions, as this pressure will insure the delivery of water to the 
fourth story of an ordinary building." 
" Various opinions have been expressed as to the quantity 
of water needed daily for a soldier on the march. In France 
* J. N. E. Water-works Assoc., xviii. i. 
t Engineering Record, June, 1904, paper 732. 
J J. Am. Water-works Assoc., March, 1915. QUANTITY OF PER CAPITA DAILY SUPPLY 
457 
it is generally estimated that each man should be provided while 
in camp with a daily supply of 30 litres (8 gallons), with an extra 
allowance of 5 litres for mounted men. The German authorities 
place the daily requirements at 50 litres (13I gallons), while 
in Austria 40 litres (io| gallons) are held to be necessary. In 
Great Britain the official barrack allowance is 24 U. S. gallons per 
man or horse, and 12 gallons for each child. Naturally, an army 
in the field is in a different position from a soldier in barracks." * 
East Indian troops employed during the war in Europe 
(1915) are found to require much water. They bathe 
before each time of prayer, and, as they pray five times daily, 
the water used while in the hospital amounts to 84 U. S. gallons 
per capita.! 
* " Water and Water Engineering." 
f J. Roy. San. Inst., 36: 409. CHAPTER XI 
ACTION OF WATER UPON METALS: TANKS, PIPES, 
CONDUITS, BOILERS, ETC. 
Lead. Many waters, especially those that are soft, and par- 
ticularly water that has been distilled, will act on the common 
metals with varying degrees of intensity, lead, zinc and iron being 
those usually attacked. Because of the serious effect upon health 
resulting from the presence of lead in drinking water that 
metal deserves particular consideration; all the more so as lead 
is a cumulative poison. 
The Massachusetts State Board of Health enumerates the 
symptoms of lead poisoning as some or all of the following: 
Anaemia, constipation, indigestion, loss of appetite, thirst, 
metallic taste, abdominal pain, colic, 11 drop-wrist," blue line 
at margin of gums, lead in the urine.* 
Dr. Hunter, describing the effects of the epidemic at Pudsey, 
says: "Anaemia and debility were the most common symp- 
toms. Patients nearly always complained that they felt as 
if they would sink down from weakness, and that the least 
exertion would make them sweat freely. The majority had the 
blue gum line so characteristic of lead poisoning. Colic was 
a common symptom. Paralysis was not common, but there 
were five or six cases of almost general paralysis and in these 
cases ' drop-wrist ' was included. The amount of lead found 
in the waters producing these effects varied from .01 to 1 grain 
or more per imperial gallon.f (.143 to 14.3 per million.) " 
There is some difference of opinion among the authorities 
as to the amounts of contained lead required to condemn a 
water, but all are agreed that even small quantities should be 
* Report of Mass. State Board Health, 1898, xxxiii. 
t Thresh, "Examination of Water," p. 88. 
458 ACTION OF WATER UPON METALS: TANKS, PIPES 
459 
narrowly watched. Thus, the Massachusetts reports note 
that one-half part per million has caused serious injury.* Haines 
holds that .1 grain per U. S. gallon (1.71 per million) should cause 
a water to be rejected.f 
Whitelegge believes that11 No water should be used for drink- 
ing which contains more that one part of lead per million, and 
any trace, however minute, indicates danger." J 
To quote Dr. Summerville in his paper in " Water 
11 Lead to the extent of .25 part per million is sufficient to 
condemn a potable water." 
That sundry public waters contain enough lead to prevent 
their acceptance by at least some of the standards laid down 
is shown by the fact that a few years ago it was reported that 
sixty-three cities of Massachusetts possessed public water 
supplies which contained lead in amounts varying from 85.46 
to .032 per million. In four of these cities where lead poisoning 
was pronounced the average amount of the metal present dur- 
ing ordinary day time use was one part or over per million. 
Occasional instances of " plumbism " were noticed in other 
towns and doubtless other mild or unrecognized cases occurred 
elsewhere. § 
In the 31st annual report of the London local government 
Board,|[ peaty, moorland waters are shown to be especially 
plumbo-solvent, to a degree chiefly governed by the amount of 
acidity present, and experiments show that such acidity is due, 
at least in part, to acid-form bacteria residing in the peat. 
That all peaty waters act on lead must not be inferred, as 
some very brown ones, notably from New Jersey, are without 
such action. Certain organic matters even serve as protective 
agents. 
The London report already quoted calls attention to the 
fact that moorland streams are highest in acidity during wet 
weather because of drainage from peaty surface sources and 
* Mass. State Board of Health, 1898, xxxii. 
f J. Fk. Inst., Nov., 1890. 
I Hygiene and Public Health. 
§ Mass. State Board of Health, 1898, page 543. 
|| 1901-02 Supplement on Lead Poisoning and Water Supply, vol. 2, page 426. 460 
WATER-SUPPLY 
less so during periods of no rain on account of their supply at 
such seasons coming from the flow of springs. 
In this connection it may be noted that H. W. Clark observed 
that carbonic acid in a soft water was the main factor that caused 
lead to be taken into solution by the waters from Massachusetts.* 
It is by no means new to distinguish between the " solu- 
tion " of lead and that " erosion " of the metal which some 
waters exercise whereby insoluble lead salts are formed with 
appreciable increase in the turbidity of the water. 
Such classification of the action upon lead has been developed 
with great care in the report of the London Local Government 
Board. 
For our purposes it will suffice to note that " erosion " 
does not occur in the absence of oxygen, and we are also to 
remember that from the sanitarian's point of view " erosion " 
may be fully as objectionable as " solution " if no opportu- 
nity for clarification be furnished. In fact, the former may 
readily be the greater evil of the two, because of its involving 
the possibility of the ingestion of large quantities of lead salts 
held in suspension. 
In a general way it may be said that soft waters attack lead 
and hard waters protect it, but this rule is not without numer- 
ous exceptions. Waters of acid reaction take lead into solution, 
while those of neutral or alkaline character hold the basic 
hydroxide or basic carbonate in suspension. As the latter 
class of waters often attack lead quite vigorously, the quantity 
of lead actually imbibed with an unfiltered water of this type 
may be considerably larger than in the case of a water where 
the lead is in solution. 
Instances are on record of lead pipes having been in use 
during many years without having been acted upon by the 
water passing through them. Thus Fischer cites a case where 
the pipes had served over 200 years without action. An 
interior incrustation on a lead pipe which had been in use for 
conveying water at Andernach during a period of 300 years 
was found to consist of: 
* Engineering News, Dec. i, 1904. ACTION OF WATER UPON METALS: TANKS, PIPES 
461 
PbO 73-962 
BiO3 0.453 
CdO 0.120 
CuO 0.323 
Fe2O3 1.552 
A12O3 1-035 
CaO 1.095 
MgO 0.283 
P2Os 8.446 
CO2 1. no 
Cl - 1.254 
Organic matter 0.388 
Si2O and clay 4-399 
Water 6.141 
100.561 
The organic matter was said to have been caused by eels 
which had been formerly employed to clear the pipe from 
material that had clogged it.* 
A marked difference commonly exists between the action of 
the same water upon new, bright lead and upon that which is 
dull from exposure, i.e., " old lead." 
Thus the writer found the following amounts of the metal 
(partly dissolved and partly suspended) in city rain-water which 
had been stored three and a half months in contact with lead 
surfaces of the above description. 
Old lead 3.65 parts per million 
New lead 58-10 " 11 
The important lesson derived from this is that lead-lined 
tanks for storage of rain-water, such as are often seen in the 
country, may grossly contaminate the water, especially while 
new. If lead cisterns or storage-tanks be deemed necessary they 
should always be carefully painted on the inside with a good 
carbon (non-metallic) paint, and should be frequently inspected. 
* J. Chem. Soc., xxxviii, 198. 462 
WATER-SUPPLY 
As has been mentioned before, there is danger in having 
a suction-pipe of unprotected lead leading to the bottom of a 
domestic well or a rain-water cistern. 
As already said, all waters do not act upon lead, and some 
very quickly form upon the metal a permanent protective coat- 
ing. Carbonate of calcium is efficient in protecting lead from 
attack. 
Crookes, Odling, and Tidy have shown the great protecting 
power of calcium silicate, and their belief is that water becomes 
lead-proof when the contained silica amounts to about 7 parts 
per million.* 
Where circumstances permit, an excellent method of checking 
the lead-dissolving powers of a soft water intended for city 
supply is to admit to the reservoir or mains a suitable quantity 
of temporarily hard spring-water. The amount of such spring- 
water required would depend upon its composition, but would 
be usually very small. 
Another remedy is to allow the entering water to flow over 
a bed of marble or limestone, or else add to it a very fine suspen- 
sion of chalk, as is done at Sheffield, England. 
The above procedure is available for treatment of peaty, acid 
waters of the moorlands; but when the plumbo-solvency is 
due to the carbon dioxide so frequently carried by ground 
waters, greater success will follow the use of simple aeration by 
means of a spray nozzle, as was demonstrated at Lowell, Mass. 
The report says: f 
" The carbonic acid can be reduced from 45 parts per million 
in the raw water to about 3.3 parts per million in the aerated 
water, and with this reduction the effect of the water on lead 
pipes is so far lessened as to tender the supply safe. 
" The removal of carbonic acid by aeration has a great 
advantage over the lime treatment in that it does not increase 
the hardness, and further, as indicated by the tests, aeration 
is more effective than the lime treatment in reducing the cor- 
* J. Soc. Chern. Ind., vii. 15. 
f Report on " Improvement of the Water Supply of the City of Lowell, Mass.," 
made in June, 1914, by F. A. Barbour, James H. Carmichael and Robert J. Thomas. ACTION OF WATER UPON METALS: TANKS, PIPES 
463 
rosive action when an equal amount of carbonic acid is left 
in the treated water. This result is not what might be ex- 
pected, because by aeration the dissolved oxygen, which has 
generally been considered a contributory factor in corrosion, 
is increased, but the experiments clearly indicate that, with 
an equal content of carbonic acid, the aerated water takes up 
less than the lime-treated water." 
That carbon dioxide in water will cause it to dissolve lead is 
easily shown by the routine experiment of blowing the breath 
through a sample of water, immersing therein a piece of the 
bright metal for a few hours and then testing for lead as usual. 
Zinc. Piping water in tubes of galvanized iron is common, 
and as zinc is often as easily attacked as lead it is pertinent to ask 
if it be equally dangerous. So far as our present experience can 
guide us toward a correct answer to this question, the reply 
must be a negative one and the following opinions are presented 
in support of such contention: 
In the Journal of the German Society of Gas and Water 
Engineers of 1887, H. Bante collected statistics to show " that 
the use of galvanized pipes should be in no way detrimental to 
health." 
Similar views are entertained by V. Ehumann, director of 
the Water Supply of Wiirtemburg.* 
According to Thresh f " There is no doubt that waters con- 
taining traces of zinc are used continuously for long periods 
without causing any obvious ill effects. The water-supply to 
a small hospital with which I was connected for some years 
always contained a trace of zinc, probably never more than half 
a grain of the carbonate per imperial gallon (7.1 parts per 
million), but I never observed any indications of its being dele- 
terious, although such effects were looked for." 
On the other hand D. H. Stacks J reports a case of wide- 
spread illness in a town which was caused by conveying artesian 
* J. Fk. Inst., Nov., 1890. 
f "Examination of Waters and Water Supplies," p. 85. 
J Engineering News, April 3, 1913. 464 
WATER-SUPPLY 
water, impregnated with hydrogen sulphide, through pipes of 
galvanized iron. The sulphide of zinc added to the water 
amounted to over thirty parts per million, and produced symp- 
toms of cramps, fainting spells and nausea. 
In the Massachusetts Board of Health report for 1900, page 
495, the following table is given showing amounts of zinc in 
sundry public supplies, the metal having been dissolved from 
pipes of galvanized iron or brass during ordinary use. The 
results are averages and are in parts per million: 
West Berlin 18.46 Lowell 0-33 
Milbury 3-08 Webster 0.28 
Newton 1.25 Sheffield 8.65 
Marblehead 0.85 Palmer 2.90 
Grafton 0.73 Beverly 2.71 
Wellesley 0.68 Fall River 0.07 
Fairhaven 0.52 
The first of the above, West Berlin, drew its water through 
four thousand feet of galvanized iron pipes. The quantity 
of metal dissolved therefrom was certainly large, but appears 
to have produced no evil results. " As far as is known the 
amount of zinc present in these waters as used is not sufficient 
to have any effect upon the health of the consumers of the 
water." 
11 The Board has investigated the question of the presence 
of zinc in drinking-water supplies where galvanized iron pipes 
are used and, except in case of the use of some ground-waters, 
containing very large amounts of free carbonic acid, which 
dissolves zinc and many other metals very freely, the amount 
of zinc found in ordinary water supplies, where galvanized 
pipes are used, is not sufficient, in the opinion of the Board, to 
give anxiety." (Massachusetts Board of Health, 1902, xliii.) 
In a private letter of more recent date the president of the 
above-mentioned board said: " If there has been any harmful 
effects of the presence of zinc in the public drinking-water of 
the State that fact would have undoubtedly been brought to 
our attention. No statement to this effect has been made, ACTION OF WATER UPON METALS: TANKS, PIPES 
465 
nor has there seemed to this board reason for suspecting serious 
danger from this source." 
As an instance of long-continued use of a water contain- 
ing much zinc, the case of Brisbane, Queensland, should be 
quoted. In that city rain-water tanks built of galvanized iron 
are found in all the houses. The water, which is in common 
use, contains about 17.1 parts per million of zinc, yet no harm- 
ful effects have been observed.* 
Unlike lead, zinc is not a cumulative poison, therefore the 
presence of the metal in very small quantities is not so objec- 
tionable. There are not a few authorities who claim that zinc 
poisoning, through the use of water, has not been proven, 
although P. F. Frankland reports such a case arising from the 
use of water from a shallow, sewage-polluted well. Waters 
from such wells were long ago shown to act quickly upon zinc.f 
In the Analyst, iv. 51, is a report of an analysis of the 
spring-water supply of Tuttendorf, Germany. The zinc present 
corresponds to .007 part of the oxide per million; the water 
has been in use a century. 
The great insolubility of those compounds of zinc com- 
monly formed by the action of water upon the metal consti- 
tutes a material safeguard. 
The writer has been unable to trace any evil effects due to 
the presence of zinc in drinking water, even when the quantity 
rose as high as 23 parts per million in a water which is in 
constant use. It might be well to add, that in the particular 
case just cited the zinc was derived from a long stretch of 
galvanized iron pipes and the amount of the metal present 
was subject to great and frequent fluctuations for reasons that 
were not apparent. 
Dr. Boardman believed that oxide of zinc, as it occurs in 
drinking-water, is absolutely harmless. He said the same of 
the carbonate. As to salts in solution, he added: " Admitting, 
then, that water which has been stored in reservoirs or drawn 
through pipes of galvanized iron always contains zinc in solu- 
* Hazen. Eng. News, April 4, 1907. 
f Rivers Pollution Commission, 6th Report. 466 
WATER-SUPPLY 
tion, in the form of one or more of its salts, the innocuity of 
those salts, in the quantities in which they occur, is attested 
by the experience and experiments of distinguished observers." 
He further said: "At least with water fit for drinking 
purposes in other respects the contained zinc salts in solution 
do not exert any deleterious effects upon the human system. 
Even if all the zinc in solution were in the form of chloride, 
the most active poison of the zinc salts, the amount would still 
be insufficient to endanger health." 
However willing most of us may be to agree with the doctor 
in his first remarks, it would be doubtful policy to follow him 
to the extent of this final statement. 
In one instance, known to the writer, a glassful of zinc 
chloride solution was taken in mistake for Hunyadi water. 
Vomiting immediately ensued and very serious illness fol- 
lowed, but the final recovery was complete. 
We know that certain waters can attack the metal to an 
objectionable degree, hence the use of galvanized iron for trans- 
mission of a water-supply should not be decided upon until 
examination has shown the water in question to be without 
material action upon the zinc coating. 
Although we are aware in a general way that softness, acidity, 
dissolved gases and the presence of much chloride or nitrate 
will tend towards metallic solvency, while alkalinity and hard- 
ness are rated as protective agents, yet it is far better to actually 
test a water with reference to its behavior toward metals than 
to attempt any prophecy of its action based upon analytical 
knowledge of what the water may contain. 
Reports recording that a water holds so many parts per 
million of lead, zinc or other metal are common enough, but 
it is rare to find advance statements of what it is capable of 
doing in the way of dissolving metals with which it may be 
brought in contact. In other words, a client who possesses a 
water-supply which is very desirable at its source is seldom 
informed of the possible damage it may sustain if conveyed 
through metallic piping. ACTION OF WATER UPON METALS: TANKS, PIPES 
467 
After the pipes have been laid and the water admitted 
to them, record is made of the result as to metallic sol- 
vency, but little is found in the nature of a prophecy ante- 
dating the outlay of capital, which prophecy, had it been 
uttered in time, might have had material bearing upon the 
investment. Again, if, as occurs in a few instances, the client 
be told that the water under examination is capable of act- 
ing upon certain metals, he is not given the information in 
such quantitative form as will enable him to .make com- 
parison between it and other waters with reference to this 
characteristic. 
Copper. Although copper is somewhat attacked by soft 
water, the result is vastly less dangerous to health than is the 
case with lead. Very small doses of copper are apparently 
not harmful to human beings, while they are destructive of cer- 
tain lower forms of life. It is the very reverse of lead in this 
particular, the latter metal being pathogenic to man and harm- 
less to micro-organisms.* 
The presence of copper in public waters does not become 
a question of practical importance except in the unlikely case of 
over " coppering " a reservoir for algae removal. 
Iron. When present, this metal is ordinarily in the water 
before it enters the distributing-mains, and is not a result of 
action upon the iron pipes. Chalybeate waters commonly 
hold the iron in solution as a carbonate, and less frequently as 
a sulphate. Inasmuch as | grain of the metal per gallon 
(4.25 parts per million) will give a distinct taste, it would be 
difficult to make a supply holding that amount popular with 
the public, even were the water not unsuited to a variety of 
uses, such as dyeing and washing. 
Action of water upon conduits of iron commonly results in 
damage to the pipe rather than to the water, either through 
corrosive attack tending to a weakening of the walls or else by 
causing a building up of interior accumulations that reduce the 
carrying capacity. The illustration shown on page 468 exhibits 
* See J. Roy. San. Inst., xxxv. 326. 468 
WATER-SUPPLY 
TUBERCLES FROM A 16-INCH WATER MAIN. (Full size.) 
WATER PIPE ALMOST COMPLETELY CLOSED BY TUBERCLES. (Full SiZe.) ACTION OF WATER UPON METALS: TANKS, PIPES 
469 
tuberculation which has almost completely blocked a .pipe 
11 inches in diameter. 
Excessive tuberculation appears to be confined to pipes 
of small diameter, the larger sizes seeming to accumulate 
layers of deposit that are comparatively thin. Neverthe- 
less whoever has observed the operation of a" scraper " upon 
some miles of a large main will recall the tons of rusty tubercle 
material removed. 
These conical masses are reddish-yellow on the surface 
and when newly removed are darker within, due to iron in 
the ferrous condition. When cut in section they exhibit con- 
centric layers of accretions and many of them show a con- 
cavity in the center of the base. Those pictured on page 468 
are from a sixteen-inch main carrying the Grafton portion of 
the Troy water-supply, and are shown in full size. 
Cutting out tuberculations by use of a scraper will, in 
small-sized pipes, increase the chances of blocking the house 
connections. Properly applied " pipe coating " is the most 
efficient remedy against these concretions. 
Inasmuch as the removal of tuberculations will increase 
the rate of subsequent pipe corrosion it becomes necessary when 
considering the advisability of cleaning large mains, to balance 
the advantage of a greater carrying capacity against the cost 
of frequent scraping. 
Macfadzean * reports that, two years after cleaning, a 
sixteen-inch main had lost 44 per cent of what it had gained in 
carrying capacity by scraping. 
Temporary deterioration of the carried water is marked in 
instances where the pipes lie empty a portion of the time, as 
when a surface pipe is drained in winter to avoid freezing. 
Iron corrodes very rapidly under such circumstances, and 
when the water is turned on again in the spring the iron oxide 
stains it for a considerable time. 
Broadly stated, highly colored swamp waters and those 
high in saline contents or in dissolved oxygen and carbon 
dioxide belong to the class which attacks iron. 
* Surveyor, 47: 772. 470 
WATER-SUPPLY 
THE MACHINE USED IN CLEANING 2O-INCH PIPE. 
N. E. W. W. Assoc., vol. xxvii. Paper by Saville on Cleaning Water Mains. ACTION OF WATER UPON METALS: TANKS, PIPES 
471 
Cast-iron pipes corrode more quickly in water containing 
an admixture of salt; as in the street-mains laid near the New 
York docks. " The life of a pipe is very short in such locality, 
and sixteen to twenty-five years is probably the limit of ser- 
vice." * 
The following from Trautwine deals with the special action 
of sea-water upon iron: 
" Genl. Pasley examined cannon and other metal from the 
wreck of the Edgar, which had been sunk in sea-water for one 
hundred and thirty-three years, and reports that the cast iron 
had generally become quite soft, and in some cases resembled 
plumbago. Some of the shot, when exposed to the air, became 
hot, and burst into many pieces. The wrought iron was not 
so much injured, except when in contact with copper or brass 
gun-metal. Neither of these last was much affected, except 
when in contact with iron." 
Some time since the writer received from Mr. W. C. Hawley, 
then Superintendent of the Atlantic City Water Department, 
a piece of corroded cast iron which had been " blown out " 
from the side of a 12-inch force-main. No analysis was made of 
it, but in appearance it resembled compact clay, yielded to the 
knife like hard clay, was easily powdered in a porcelain mortar, 
and had a specific gravity of 2.28. 
The following is from Mr. Hawley's letter: " The pipe has 
been laid about nineteen years under a salt meadow of deep 
black mud. Two separate layers of sound sod lie below the 
present surface grass. 
" The action of this mud on steel or wrought iron is ex- 
tremely severe. Wrought iron after fifteen or eighteen months 
looked as though it had been placed in strong acid, the fiber 
being brought out distinctly. The deterioration of the force- 
main ranged from one-half to the full thickness of the metal." 
The rusting of iron is a complex process and scientists are 
very far from being agreed upon just what takes place during 
the procedure. Whitney's electrolytic theory has as many 
*Jour. Am. Water-works Assoc., xii. 27. 472 
WATER-SUPPLY 
supporters as any other, but until the specialists have reached 
some final conclusion the water engineer may be allowed to 
retain the chemical conception which has been the longest in 
vogue. 
2Fe+4CO2 + 2H2O+O2 = 2Fe(HCOs)2 
4Fe(HCOs)2+O2 = 2Fe2O3d-4H2Od-8CO2. 
The above equation are to considered as suggestive only. 
Without doubt the actual reactions are much more compli- 
cated. It will be noted that the carbon dioxide required to 
start the process is subsequently released and may repeat its 
action indefinitely. 
In the matter of resisting corrosion cast iron is apparently 
superior to steel or wrought iron provided the " foundry skin " 
be not removed. 
Thwaites * showed this experimentally at some length, 
thereby substantiating the generally accepted view. Of course 
ordinary commercial brands of the metals are here referred to, 
for it is well known that pure iron is highly resistant to corro- 
sion. 
11 Many ancient monuments and relics in wrought iron still 
in a good state of preservation indicate that the wrought iron 
of those days was decidedly superior to the modern commercial 
material. 
" The iron pillar at Delhi is one of the most interesting relics 
of ancient India, and is said to date back from the year 912 
b.c. This pillar, 50 feet in height, and about 16 inches in diam- 
eter, is built up of blooms, weighing about 50 pounds each, 
welded together. The pillar is remarkably free from rust, 
though the climate of Delhi is not particularly favorable for 
the preservation of iron exposed to the atmosphere. The sur- 
face of the iron has a bronze-like appearance, and one of the 
suggestions which has been put forward to account for its re- 
markable state of preservation is that the iron was forged on a 
siliceous stone anvil, and that some of the silica of the stone 
entered into combination with the iron, as it is known that 
* Water and Water Engineering, June 15, 1915. ACTION OF WATER UPON METALS: TANKS, PIPES 
473 
silica has a marked influence in resisting corrosion when alloyed 
with iron." 
As to what should be considered the normal life of good 
quality cast-iron water pipe, knowledge of the locality where 
the pipe is to be laid must be had before an estimate can be 
risked. As already noted, sea-water and salt marshes are 
conducive to short periods of service, while in good soil the 
life of such pipe is so long as to exceed our present period of 
experience. 
Boilers may be affected by water in several ways, namely, 
through the corrosive action of what the water may hold in 
solution, or, indirectly, through the secondary evils resulting 
from foaming and scale formation. 
Tendency toward foaming is caused by the presence of 
suspended matter of organic character, or even inorganic 
materials if they be finely divided, also by salts of the alkalies 
in solution and by any substances which increase the viscosity 
of the water. Henderson says: * 
"If a water contains over 50 grains of foaming matter 
(soluble salts) to the gallon (850 parts per million), it is sure to 
give trouble in a locomotive." 
" One hundred grains per gallon is made the limit of con- 
centration in the locomotive.f This means that the amount 
in the original water must be much lower. Forty to fifty grains 
per gallon is the limit for waters to be used in locomotives. Much 
higher concentration can be obtained in the stationary boiler 
without experiencing any difficulty." 
Reduction of the concentration, at the same time removal 
of sludge, by blowing out the boiler, would seem to be the logi- 
cal treatment for foaming, although it entails expense through 
loss of heat. 
Any free acid is objectionable in a boiler-water, even car- 
bonic acid, if the quantity be large. In fact, altogether too 
* Engineering News, 50: 280. 
t Report of Committee, Am. Ry. Eng. & Maintenance of Way Association 
Bulletin 83, p. 45. 474 
WATER-SUPPLY 
little attention is paid to the deleterious action upon the metal 
by the gases which the water may hold in solution. 
Carbon dioxide and atmospheric oxygen act upon iron at 
high temperature with decided energy, and such action may 
occur at a considerable distance from the boiler itself. The 
writer has observed serious effects, due to dissolved gases, 
upon the steam pipes many feet from the boiler. 
It should be remembered that inasmuch as oxygen is more 
soluble in water than is nitrogen the " dissolved air " is much 
richer in oxygen than the atmospheric ratio calls for. 
Sulphuric acid, so commonly present in mine-water, is 
highly objectionable. It is formed from the decomposition 
of iron pyrites: 
2FeS2 + 7O2 + 2H2O = 2FeSO4+2H2SO4. 
Sundry streams of western Pennsylvania are so polluted 
by mine drainage as to become distinctly acid, especially dur- 
ing periods of low rainfall. Thus the average acidity of the 
Youghiogheny river for the year ending Sept. 1, 1907, amounted 
to 22 parts per million, measured as sulphuric acid.* The 
effect of such a water upon plumbing fixtures is very severe. 
The liberation of free fatty acid by the steam acting under 
pressure upon lubricating oils of animal or vegetable origin 
is a common cause of corrosion. The widely known " Tilghman 
process " illustrates such action, for instance margarine breaks 
into glycerine and margaric acid when so heated. 
C3H5(C17H33O2)3+3H2O = C3H5(OH)3+3H(C17H33O2). 
For this reason it is preferable to employ a cylinder-oil of hydro- 
carbon character. 
Water containing magnesium chloride is especially to be 
avoided for boiler uses, because the salt decomposes at the 
high temperature attained, with production of free hydrochloric 
acid according to the equation 
MgCl2 + 2H2O = Mg(OH) 2+2HCI. 
* Water Supply Paper 236, U. S. Geol. Sur. ACTION OF WATER UPON METALS: TANKS, PIPES 
475 
This acid being readily carried over in the steam, the damage 
that it works is not confined to the boiler alone. 
It being known that ammonium chloride will prevent the 
decomposition of magnesium chloride, during evaporation, by 
forming therewith a stable double chloride, A. H. Allen sug- 
gests that the sodium chloride of sea-water acts in a similar 
way for the protection of marine boilers from the magnesium 
chloride found in sea-water. His remedy for stationary boilers 
compelled to use magnesium waters is to add common salt to 
the feed-water.* 
In view of the bad effects of magnesium chloride upon 
boilers, he further contends that it should appear in the analysis 
to the fullest extent compatible with the total amounts of 
chlorine and magnesium. 
Water strongly alkaline with sodium salts, as is found in 
certain sections of the West, is also corrosive; for instance, such 
a water as that from Bitter Creek, Wyoming, which contains: 
Per Million. 
Calcium carbonate 13.1 
Silica (clay) 3.9 
Calcium sulphate trace 
Sodium sulphate 431.0 
Sodium carbonate 843.1 
Sodium chloride 96.3 
Silica (in solution) 8.6 
Such a water could be purified for boiler purposes by the 
use of barium chloride; but another, and possibly cheaper, 
method under the circumstances is that employed by Mr. A. 
Pennell. He writes to the author: " Calcium sulphate is 
added, which forms sodium sulphate and precipitates cal- 
cium carbonate. A further dose of gypsum is then added, 
and the water is heated to 2000 F., whereupon glauberite 
(Na2SO4.CaSOi) precipitates as semi-transparent crystals. All 
does not precipitate at this temperature, but the rest falls at 
boiler temperature and is blown off at intervals." 
*J. Soc. Chem. Ind., vii. 800. 476 
WATER-SUPPLY 
Boiler-scale may be classified as of two general kinds: first, 
that which is friable and mud-forming, such as is caused by the 
employment of temporarily hard water; and, second, a hard, 
compact, and adherent form, arising from the use of water of 
permanent hardness. 
The hard adherent form is much the more objectionable, 
as a mud deposit is readily removed. The cause of the deposit 
of the calcium sulphate, which forms the compact scale, is found 
BOILER TUBE NEARLY BLOCKED BY HARD SCALE. Full Size. 
in the insolubility of that salt at the high temperature attained 
in the boiler. The curve of solubility shown on page 477 is 
seen to closely approach the zero line at a temperature of 1500 C. 
The writer possesses some hard, dense sulphate scale, of 
two inches in thickness, which was taken from the boilers of 
the steamer Tybee. 
The heat arrested by a thick deposit of scale is ruinously 
wasteful of fuel. 
Thus Rankine estimates * that the conductivity for heat of 
* " Water Purification," Rideal, p. 201. ACTION OF WATER UPON METALS: TANKS, PIPES 
477 
Dotted line hypothetical- 
Numbers on curve are the 
fractions of parts held in 
solution by IOO parts of, 
water at the temperature 
noted on the horizontal line- 
CURVE OF SOLUBILITY OE CALCIUM SULPHATE. 478 
WATER-SUPPLY 
calcium carbonate is eighteen times, and of calcium sulphate 
is fifteen times, less than iron. 
He holds that 
| inch scale requires 16 per cent additional fuel 
ICC CC Cl Cl Cl ll 
5° 
" 11 11 150 11 11 11 
As a result of tests upon a locomotive boiler it was found 
that scaling to the depth of | inch caused a loss of steaming 
efficiency equal to 10.5 per cent.* 
Outside of waste of fuel, the damage to the boiler-metal 
resulting from the high temperature to which it must be heated 
in order to overcome the scale resistance must also be taken 
into consideration. 
The deposits of the carbonates held in solution in tempo- 
rarily hard water is caused by the escape of the solvent carbon 
dioxide upon the rise in temperature of the boiler water. 
Should means other than the elevation of temperature be 
employed for the removal of the carbon dioxide in solution, the 
precipitation of the dissolved carbonates would take place with 
equal certainty. 
Thus many years ago Dr. Clark patented a process, which 
still bears his name, for removing the carbon dioxide by the use 
of lime-water, according to the equation 
CO2 + Ca(OH)2 = CaCO3+H2O. 
The calcium carbonate formed by the equation precipitates 
and along with it also fall the calcium and magnesium car- 
bonates originally held in solution in the water. 
Bartow and Scholl have shown that magnesium is without 
value in the softening of water and that it increases the amount 
of sludge to be handled without corresponding benefit. Con- 
sequently a magnesium-free lime should be procured for the 
softening process, f 
In America the " Clark process " for softening temporarily 
* J. Am. Chem. Soc, xxviii. 640. 
t J. Indust, and Engr. Chem. 6: 189. ACTION OF WATER UPON METALS: TANKS, PIPES 
479 
hard waters is not very frequently resorted to, because our 
waters are commonly fairly soft, or else are permanently hard, 
a form of hardness for which the process is not suited. 
In England, however, where chalk deposits are so plenty, 
this method of purification is more often seen, and even on so 
large a scale as that required for a city supply. At Southamp- 
ton the water for 63,500 persons comes from a large well in the 
chalk, sunk in 1888; and it is softened by a " Clark process " 
plant of a capacity of 2,000,000 gallons daily. The water 
receives a charge of 10 per cent of its volume of lime-water in 
a mixer and is then discharged into a softening cistern 38 by 
23 by 3 feet. After partial precipitation, the milky water 
passes to perforated filter-plates covered with cloth. The cost 
of this plant was about $50,000. It produces tons of pre- 
cipitate daily, and uses up | ton of lime for the purpose.* 
For boiler purposes the expensive filter-presses would not 
be warranted, and simple settling-tanks should be substituted. 
Care should be taken to avoid the introduction of more 
lime-water than the reaction calls for, as a large excess would 
of itself cause a boiler incrustation. The brown precipitate 
produced by pouring a solution of silver nitrate into lime-water 
is a convenient indicator for use with the Clark process. As 
soon as the said brown precipitate appears, in a sample of the 
treated water, upon addition of a few drops of silver nitrate 
solution, the further introduction of lime-water should cease. 
The softening of permanently hard water may be accom- 
plished by the addition of a solution of sodium carbonate, 
which causes a precipitation of insoluble'calcium carbonate: 
CaSO4+Na2CO3 = CaCO3+Na2SO4. 
Whatever method is used the softening process should 
be applied outside the boiler, as by doing so the deposited sludge 
does not have to be subsequently blown off. Moreover if pre- 
liminary heating of the feed-water be undertaken it can be so 
conducted as to remove from solution any corroding gases. 
As a matter of practice, however, the boiler itself is in many 
* Engineering, March n, 1892; see also Engineering News, April 16, 1892. 480 
WATER-SUPPLY 
instances forced to serve as a reaction chamber and the dose 
of chemicals is added directly. 
Complete softening is not contemplated, because of additional 
expense and for the reason that a slight film of scale in a boiler 
is rather beneficial than otherwise. 
As showing the views held by those who make use of boiler 
waters on a large scale, the following extracts from letters, 
written to the author during 1915, may be of interest. They 
are from Engineers of Tests connected with some of the largest 
and most important railroads in the United States: 
No. 1. " About the best water we have analyses from 
120 to 170 parts per million of hardness. We would, of course, 
prefer water with not over 85 parts per million of hardness and 
consider a water containing 255 to 340 as poor, and above that 
as bad. 
" The limit of hardness in water which can be successfully 
treated depends largely upon local conditions and the chemicals 
with which the water is treated. Lime and soda-ash are the 
ordinary chemicals used on this railroad, although more expen- 
sive chemicals, such as barium compounds, are used in some 
places. I do not believe railroads would resort to the use of 
the expensive chemicals in question unless the water already 
carried 340 to 510 parts per million of sodium sulphate. 
" The degree of hardness which can be successfully treated 
will vary with the importance of the water station. At some 
outlying points where comparatively little water is used it can 
be treated without any undue trouble from foaming in engines 
when the hardness is as high as 1275 parts per million. 
"It would not be advisable, however, to attempt treatment 
of water of this character at an important station where all 
engines take a large amount of water." 
No. 2. "I consider a water as good for boiler purposes 
if it has a hardness of not over 170 parts per million, provided 
not more than 50 to 70 parts per million of this hardness is 
due to calcium and magnesium sulphates. ACTION OF WATER UPON METALS: TANKS, PIPES 
481 
" I would consider a water having a total hardness of 680 
parts per million and over, of which 255 parts per million is 
classified as permanent hardness, as too bad for boiler pur- 
poses. I have found from practical experience that a water 
containing over 500 parts per million of permanent hardness 
cannot satisfactorily be treated by the soda-lime process, as 
a water so treated as to remove all the incrusting sulphates would 
cause difficulty from foaming in locomotive boilers." 
No. 3. "A good water would be one with a hardness of 
less than 170 parts per million; a poor water would have a hard- 
ness of from 340 to 510 parts per million. A water above 
680 parts per million would be considered too bad for use and we 
would not care to go to the expense of treating water of this 
character, as above that figure it is not profitable." 
No. 4. " In our experience a good water contains less 
than 85 parts per million of total residue, is free from acid salts, 
magnesium chloride, and other constituents which might be 
corrosive and it should contain a small amount of carbonates. 
" The point at which it becomes profitable to soften water 
depends very largely on the type of boiler in which it is used." 
In some of our boilers " we are actually treating water that 
contains only 100 to 120 parts per million of total residue and 
the treatment is undoubtedly profitable. In most cases we 
have succeeded in successfully treating boiler water containing 
up to 510 parts per million total residue. Some of our sub- 
sidiary lines are treating water containing from 510 to approx- 
imately 1020 parts per million of total residue with some degree 
of success. Water containing a higher residue than this is not 
used except in case of absolute necessity." 
No. 5. " In regard to the rating of boiler waters on this 
system, we have a scheme which is shown as follows." 
INCRUSTING RATING (PARTS PER MILLION) 
1. Very good. Water having sodium carbonate and hard- 
ness less than 350. 
2. Good. Water having sodium carbonate and hardness 482 
WATER-SUPPLY 
greater than 350. Water having sulphate hardness less than 
50 and total hardness less than 200. 
3. Fair. Water with sulphate hardness between 50 and 100 
and total hardness less than 300, or total hardness between 
200 and 300. 
4. Bad. Water with sulphate hardness between 100 and 150 
and total hardness less than 500, or total hardness between 
300 and 500. 
5. Very bad. Water with sulphate hardness greater than 
150 or total hardness greater than 500. 
FOAMING RATING (PARTS PER MILLION) 
A. Very good. Alkali salts less than 70. 
B. Good. Alkali salts between 70 and 150. 
C. Fair. Alkali salts between 150 and 250. 
D. Bad. Alkali salts between 250 and 400. 
E. Very bad. Alkali salts over 400. 
No. 6. Classification for boiler use (parts per million). 
Good. Water with o to 100 scale-forming solids. 
Fair. Water with 100 to 170 scale-forming solids. 
Poor. Water with 170 to 240 scale-forming solids. 
Waters for boiler use (in Connecticut) have been classified 
by G. H. Seyms as follows, based upon the total solids stated 
in parts per million: * 
Good, less than 250. 
Fair, between 250 and 500. 
Unfit, over 500. 
Scale from sea-water consists mainly of calcium sulphate 
and magnesium hydroxide; in fact, as Driffield has shown, 
magnesium occurs in these deposits as hydroxide, although 
precipitated as carbonate, the conversion to the former being 
accomplished by the high temperature of the boiler.f 
* Water Supply Paper 232, p. 16g. 
f J. Soc. Chem. Ind., vi. 178. ACTION OF WATER UPON METALS: TANKS, PIPES 
483 
Such a change is more likely where the scale is in touch 
with highly heated metal, and it may be represented by the 
equation 
MgCO3 + 2H2O = Mg(OH)2 + CO2+H2O. 
A very large number of boiler-scale " preventives " and 
" eradicators " have been placed upon the market which are 
peculiar for nothing, as Professor C. F. Chandler has well said, 
except their high price. Such as have any value whatever 
may be duplicated, at very little expense, out of quite common 
materials. 
Unfortunately, many of these preparations are perfectly 
inert, and not a few are positively harmful. In the latter class, 
for instance, the writer has found such material as acid sodium 
sulphate colored with logwood. Such a preparation acts upon 
metals with half the intensity of pure' sulphuric acid, and its 
continued use must surely work injury to the boiler. A large 
class of these " preventives " aim not at the actual prevention 
of a deposit, but rather seek to alter its physical character. 
Thus many of them are of a mucilaginous order, and they 
so envelope the precipitating particles of mineral matter as 
to prevent their mutual coherence. A further action of such 
of the compounds as contain insoluble material like sawdust 
is to provide separated nuclei, about which crystallization 
of the scale-forming salts may occur. In the first instance, 
such an increase in the viscosity of the water may follow as 
to cause serious frothing or " priming," and in the second 
there is additional danger of getting solid substances carried 
over into the moving parts. It is very questionable if as 
desirable results can be obtained at reasonable price by the 
employment of any o^ the " eradicators " as may be had by 
the use of ordinary sodium carbonate added in proper 
quantity. 
Good results are obtainable from the use of an iron-zinc 
couple, secured by attaching plates of zinc to the boiler brac- 
ings. Protection of the iron follows at the expense of the 
zinc plates. 484 
WATER-SUPPLY 
The excess lime method of softening and disinfecting water, 
proposed in 1912 by Dr. A. C. Houston of the Metropolitan 
Water Board (London), is a modification of the Clark process, 
and is fully described in the Eighth Research Report. Briefly 
stated, the process consists in dosing a fraction of the volume 
of water to be treated, say 75 per cent, with the full quantity 
of lime intended for all of the water and then, after an interval 
of twenty-four hours, adding the remaining 25 per cent of the 
water. The heavy overdose of lime used in the first fraction 
will, during the twenty-four hours interval, kill objectionable 
organisms that may be present and will also furnish the proper 
softening dose for both portions when they are brought together. 
The smaller part is separately sterilized by storage or other 
means before the mixture is made. The method has not been 
tried in America. The claim made in its favor is the lowering 
of cost due to the very cheap means taken to sterilize three- 
quarters of the quantity of water handled. The process is 
preliminary to filtration. 
Houston says: "When 1 part of quicklime (about 75 per 
cent CaO) is added to 5000 parts of raw Thames water about 
0.007 per cent free CaO is left in the mixture and this excess is 
sufficient to kill B. coli in from five to twenty-four hours. To 
neutralize the excess CaO, not less than 25 per cent of stored 
water must subsequently be added." 
Industrial waters should be in most cases " soft " in char- 
acter, although a hardness due to calcium sulphate is pref- 
erable for the use of breweries that make light-colored beer. 
The laundry interests in particular require a soft water not 
only to lessen the expense due to soap destruction (which 
amounts to ten cents per million gallons of water for each 
additional part per million of hardness, as shown by Whipple), 
but also because the " lime soaps " formed by the use of hard 
water are likely to enter the body of the cloth in the shape of 
small lumps which are difficult to remove. The laundries 
are also desirous of avoiding water containing iron as its use 
causes rust spots upon white goods. ACTION OF WATER UPON METALS: TANKS, PIPES 
485 
For paper making clearness of the water employed is a prime 
requisite. Iron, even in solution, is objectionable for prac- 
tically the same reason as in the laundry business; it spots 
the finished product. As differing from the laundry-man, 
however, the paper-maker prefers a somewhat permanently 
hard water to a very soft one for the reason that it is less waste- 
ful of the calcium sulphate " loading " that is frequently 
employed. 
In dyeing processes the absence of turbidity and iron is the 
chief demand, although softness is also important. This is 
also true for the tanning industry. 
Permutit, an artificial zeolite, has of recent date been intro- 
duced as a softener of water. In general appearance it resem- 
bles ordinary filter-sand, but the grains possess consider- 
able porosity. It is made by fusing together, in suitable pro- 
portions, quartz, clay and soda-ash, the result being represented 
by the formula Na(A12Si2O8). When water carrying either 
CaSO4 or CaCOs filters through a bed of this material the 
calcium and sodium exchange places, the water becoming soft- 
ened as per the equations: 
2Na(Al2Si208) + CaSO4 = Ca(Al2Si2O8)2+Na2SO4 
2Na(A12Si2O8) TCaCOa = Ca(A12Si2O8)2TNa2CO3. 
The calcium permutit so formed will slowly react with a 
io per cent solution of common salt in the following manner: 
Ca(A12Si2O8)2 + 2NaCl = 2Na(Al2Si2O8) + CaC12. 
Hence, if, after using the apparatus all day for water soft- 
ening the permutit bed be allowed to rest during the night 
covered with salt solution, regeneration of the filter becomes 
completed by morning and, after draining off the solution of 
calcium chloride to the sewer, the plant is again ready for its 
work of water softening during the day. Thus the upkeep of 
the process is reduced to supplying the daily amount of common 
salt used. For laundry purposes the result is favorable. In 
one small public laundry in Troy, where a Permutit plant was 486 
WATER-SUPPLY 
installed at a cost of $1200, the manager finds a saving of $15 
weekly in soap, at an outlay of $2 weekly for the salt required. 
The soap hardness of the raw water used averages about 40 
parts per million. 
The use of manganese permutit possesses some advan- 
tages as a method for manganese removal. 
Representing sodium and manganese permutits by the 
shortened formulae Na2P and MnP, the reactions for making 
manganese permutit and its use for removal of manganese 
from the water would stand: 
Na2P+MnCl2 = MnP + 2NaCl 
MnP + 2NaMnO4 = Na2P+MnO - Mn2O?. 
Na2P+MnO-Mn2O7 + 2Mn(HCO3)2 
= Na2P + 5MnO2+4CO2 + 2H2O. 
The MnO2 formed is removed daily by a back-wash which 
requires 1 per cent of the treated water.* 
For boiler use the Permutit process solves the question of 
scale formation, but introduces the possibility of " foaming " 
if the hardness of the raw water be high, on account of the quan- 
tity of sodium salts added to the water. In this particular it 
does not differ from the older method of softening sulphate 
waters by use of soda-ash. In either case a water may be too 
hard to permit of economic softening. 
Grease balls sometimes occur in boilers and are formed by 
oil or grease in the water uniting with precipitated calcium 
and magnesium salts also present. (See page 487.) 
The resulting insoluble " soaps " become rolled together 
by motion of the water until balls of considerable size are formed. 
The one here shown is from the Traveler''s Standard, III: 205, 
and it is stated to have measured three inches in diameter. 
Red water, a caption now often seen, describes the un- 
sightly turbidity of iron oxide that follows heavy corrosion 
* See J. Indust, and Engr. Chem., 8: 160. ACTION OF WATER UPON METALS: TANKS, PIPES 
487 
of uncoated hot-water service-pipes, and those of a cheap 
brand of galvanized iron. Nor is the corrosion confined to 
hot-water fines, although more commonly occurring in them 
than in pipes carrying cold water. When water is low in 
alkalinity and high in both dissolved oxygen and carbon dioxide, 
" red water " is likely to occur, if poorly coated iron pipes be 
installed in the hot-water system. Its occurrence seems to 
" GREASE BALL." 
vary with the seasons, being more frequently noticed during 
the colder months, when dissolved gases are carried by the water 
in greater quantity. 
Mechanical filtration has been blamed for the appearance 
of red water, the claim having been made that the alum used 
in the process causes corrosion by virtue of its acidity. Of 
course an overdose of alum, sufficiently great to produce acid- 
ity, would certainly attack iron, but in a modern, well-managed 
filter-plant no such overdosing is permitted to occur. 
It may well be that the clear water furnished by the filter 
has had removed from it certain organic constituents that before 
filtration acted as protective agents to the iron surfaces and 488 
WATER-SUPPLY 
thus allowed the corroding effect of a soft water to become 
more evident. 
When filtering a water likely to cause corrosion, if the 
alkalinity should be found to be near the amount required 
to react with the dose of alum used, it should be raised by the 
addition of lime or soda-ash to a point nearly or quite double 
that value (7.5 parts per million alkalinity per grain of alum 
used in each gallon of water), say to. 15 parts per million so that 
undue softness, with consequent " red-water " troubles may be 
avoided. 
Thorough protection of the iron piping should be secured 
by a well applied zinc coating or, better still, the hot-water 
line may be constructed of brass, or, if the pressure permit, of 
lead. 
If arrangements can be made whereby the water can be 
heated in an open heater before admission to the boiler, the con- 
sequent removal of dissolved gas will prevent corrosion in 
either the boiler or the service pipes. 
R. S. Weston finds that removal of carbon dioxide by aera- 
tion, although only partial, will do much towards preventing 
the corrosion of iron. 
Removal of the gas by means of lime or soda-ash would also 
lead to the same result.* 
The actual bulk of the material formed by pipe corrosion 
is often very large. Whipple, who has measured it, says:f 
11 It hardly seemed credible that so much iron rust could be 
derived from the pipe itself, but experiments showed that each 
volume of metallic iron was capable of producing upwards 
of 10 volumes of iron rust in a moist condition, and calculations 
showed that in order to completely fill a i|-inch pipe it would 
be only necessary to have the metal of the steel reduced on an 
average by less than 27 inch, or in a i-inch pipe by inch." 
* See Report by Jackson and Hale, New York Department of Water, 1912. 
f Report to Board of Water Commissioners, Springfield, Mass., 1910. APPENDIX A 
ANALYSES OF SEA WATERS (From Stillman) 
(Parts per Million.) 
Total 
Solids. 
Ca 
Mg 
Na 
K 
NHi 
Atlantic Ocean, 410 18' N.; 36° 28' W. 
Pacific Ocean, 28° n' S.; 93° 24' W.; 
420' deep 
North Sea bet. Belgium and England. . . 
Baltic Sea 
Mediterranean Sea, off Cette 
Black Sea off Crimea 
Caspian Sea 
Dead Sea 
38,400 
35.220 
32,800 
17.710 
37.706 
1,766 
6,296 
240,483 
556.8 
475.2 
324.4 
36.3 
444-1 
130.5 
191.6 
9000 
1,198.1 
1,471.4 
1,158.2 
611.5 
1,310.4 
662.2 
409.8 
19.883 
I77I9-7 
10,232.6 
10,205.7 
5.894-1 
11,706.2 
5,512 
1.444 
47,9i8 
668.2 
633.6 
353-6 
264.3 
97-5 
139-7 
6,385-1 
18 
Al 
Fe 
Mn 
CO3 
SOi 
Cl 
Br 
Atlantic Ocean 41° 18' N.; 36° 28' W. 
Pacific Ocean 28° 11' S.; 930 24' W.; 
420z deep 
3029.8 
3826.6 
2590.2 
718.9 
2942.8 
1250.2 
1337-2 
470-9 
20,839-7 
19,321.3 
18,167.9 
10,386.4 
20,526.9 
9,574-5 
2,737-7 
154,442 
387^8 
239-4 
433-5 
5 
2176.7 
North Sea bet. Belgium and England. . 
Baltic Sea 
.... 
Mediterranean Sea off Cette 
Black Sea off Crimea 
Caspian Sea 
40. 
153-4 
2.8 
[27 . I 
67-9 
247-5 
77-3 
Dead Sea 
11.7 
25.8 
489 APPENDIX B 
RIGHTS AND DUTIES REGARDING THE POLLUTION OF 
STREAMS 
Taken from a paper by Edwin B. Goodell, on " Water 
Supply and Irrigation," published by the United States Geo- 
logical Survey:- 
i. Rights and Duties of Riparian Owners 
Every riparian owner has the right- 
i. To use the waters of streams, navigable or otherwise, 
which flow across or along his property for the ordinary purposes 
incidental to domestic life and agriculture, including grazing. 
2. To use such waters for water power, and for all kinds 
of manufacturing purposes which do not sensibly diminish the 
quantity which flows on for the use of lower proprietors, nor 
change the quality of the waters to any appreciable extent, nor 
interfere with the use of the stream, if navigable, by the public. 
3. To have such waters flow to him from the premises of 
higher proprietors not unreasonably diminished nor diverted, 
nor rendered impure by the farming or domestic uses to which 
the waters are subjected by higher proprietors. 
4. To have such waters flow to him not sensibly changed in 
quality by any manufacturing or other uses to which they may 
have been put by higher proprietors. 
5. To have such waters flow to him in their natural bed, 
unpolluted by any deposits of filth, or any other substance 
in the bed or channel previously traversed by them. But 
3, 4, and 5 do not apply to riparian owners in those States in 
which the doctrine of prior appropriation is the law. 
Conversely, it is the duty of every riparian owner- 
1. To so guard his use of the waters of streams which flow 
490 APPENDIX B 
491 
across or along his property for domestic and agricultural pur- 
poses as not unreasonably to divert, nor diminish, nor render 
impure such waters. 
2. To refrain from every use in manufacturing which will 
divert or sensible diminish the quantity of the waters which 
flow onward to the lower proprietors or render them appreciably 
different in quality. 
3. To refrain from depositing any filth or other substance 
in the bed of such streams in such a manner or to such an extent 
as will cause the waters to flow to the lower proprietors out 
of their natural bed or will in any wise pollute them or render 
them impure. 
Where the doctrine of prior appropriation is in force the 
appropriator must confine his use of the appropriated water 
to the use for which he had appropriated it and take only so 
much as is reasonably necessary to accomplish that purpose. 
He may not pollute the stream wantonly, nor by using it for 
purposes not included in his appropriation. Subject to these 
restrictions, the prior appropriator has the right to divert from 
the stream and use as much of the water as is necessary to 
accomplish the purpose for which it was appropriated. 
II. Rights and Duties of Municipal Corporations 
Considered as corporate entities, municipal corporations have 
such rights and powers only as are conferred upon them by 
statute, either expressly or by necessary implication. 
When, under due authority, they become the owners of 
lakes, reservoirs, and natural streams, they have the same 
rights to pure water, and are charged with the same duties as 
are other riparian proprietors. 
If authorized to construct a system of sewage draining 
into a stream, such authority does not exempt them (except- 
ing the State of Indiana) from the duty not to pollute the stream 
to the damage of lower proprietors. 
The rights of property owners, specified in 3, 4, and 5 above, 
are property rights and cannot be taken away from owners for 
public use except upon payment therefor of an amount deter- 492 
APPENDIX B 
mined by constitutional condemnation proceedings authorized 
by statute. 
Therefore, until municipal corporations have, by such pro- 
ceedings, acquired the right of all lower proprietors and paid 
for them, they are required in all cases to refrain from the 
pollution of streams to the same extent as private owners. 
III. Rights and Duties of the Public 
By 11 the public " is meant that indefinite number of in- 
dividuals, whether larger or smaller, who occupy as a common 
habitation a neighborhood, village, town, state, or country. 
Rights and duties which affect inhabitants of the neighborhood, 
village, town, state, or country as a whole, or a considerable but 
indefinite number of them, are called "public" rights and duties. 
The public, in this sense, aside from the right to use navi- 
gable waters for commerce, has the right to enjoy the natural 
waters and the air which passes over them, so far as life and 
health are affected by these elements, in a condition so near that 
in which nature left them that their use will not destroy nor 
threaten life nor injure health. 
And, reciprocally, the public, and each member of it, is 
charged with the duty not to pollute the natural waters upon 
which the community depends for life and health in any man- 
ner that will render the continued use of the waters, or of the 
air which passes over them, destructive of or injurious to the 
life or health of the community. 
Public Rights and Duties Enforced by Statute 
The rights and duties attempted to be expressed under 
III. have received some recognition by the Courts apart from 
statutory enactments. They have been enforced chiefly, how- 
ever, through legislation. These rights and duties have received 
full recognition, and an active effort has been made to provide 
an efficient sanction for their enforcement by the Legislature 
of all the Statutes included in Class II. and Class III., as here- 
inbefore stated. These classes include thirty-eight of the States 
and territories. APPENDIX B 
493 
These statutes, not being in derogation of common-law 
rights, have been construed as remedial statutes and not un- 
constitutional, although in some cases they may seem to inter- 
fere with prescriptive rights' No one can acquire by pre- 
scription a right to do an act which menaces public health or 
destroys public comfort. 
It will have been noticed that public opinion, as expressed 
in public laws, is steadily progressing in the direction of a full, 
complete and comprehensive enforcement of all the rights 
and duties of riparian owners, of municipal corporations, and 
of the public, as summarized above. Each advance in statu- 
tory regulation is an advance in that direction, and more 
especially in the direction of regulating and enforcing public 
rights and municipal rights and duties. 
Private owners, from time immemorial, have been active 
in protecting their riparian rights as against other private 
owners. But the effect of pollution upon public health has not, 
until a comparatively recent period, been brought prominently 
into notice. The pollution of streams by cities and private per- 
sons has, accordingly, not received the attention which it 
deserved. This state of affairs is now rapidly passing away. 
Courts have shown themselves fully alive to the existence and 
validity of public rights in that respect, and the Legislatures 
in Class III., comprising the States of Connecticut, Massa- 
chusetts, New Hampshire, New York, New Jersey, Minnesota, 
Vermont and Pennsylvania, which have come into this class by 
legislation enacted in 1905, have made enactments calculated so 
to control such pollution as eventually to prevent all danger to 
public health. APPENDIX C 
TYPHOID FEVER CONTRACTED FROM DRINKING A POL- 
LUTED WATER IS DECIDED TO BE "AN ACCIDENT" 
In 229 Federal Reporter 552 the decision of the United 
States Circuit Court of Appeals reads as follows: 
The defendant in error, the Pacific Gas and Coke Co., 
being engaged in the construction of a gas plant and having 
employed in the work a large number of men, secured from 
the Aetna Life Insurance Co., the plaintiff in error, a policy 
entitled by the latter " contractor's employers' liability policy," 
by which in consideration of certain premiums which the case 
shows the defendant in error paid, it agreed to indemnify the 
assured (within certain amounts within which the present 
case falls) against loss and expense arising or resulting from 
claims upon the assured for damages on account of bodily 
injuries or death accidentally suffered, or alleged to have been 
suffered, by an employee or employees of the assured by reason 
of the business as described and conducted at the locations 
named in the policy, with certain exceptions not applicable 
here. In the course of the work certain of the employees of 
the Gas and Coke Co. contracted typhoid fever from the water 
furnished them by the latter, on account of which that company 
was compelled to pay damages to such injured employees to 
recover the aggregate amount of which from the insurance 
company the present action was brought. And the sole point 
here presented is whether the harm so done to the workmen 
constituted a bodily injury accidentally received or suffered 
by them, within the meaning of the policy in question. 
Of course it is not and cannot be doubted that the workmen 
were bodily injured by the drinking of the water in the course 
of the work, for it contained typhoid germs, which gave them 
typhoid fever; but it is insisted on the part of the plaintiff in 
494 APPENDIX C 
495 
error that in drinking the water they were but satisfying a 
natural want, and that in doing so there was no accident about 
it. It is readily conceded, of course, that there could be no 
accident in merely drinking water; but it is just as certain 
that the men would not have drunk it had they known that 
the water contained typhoid germs. The accident consists 
in that unexpected happening. Among the definitions of the 
word " accidental," in most, if not in all, of the dictionaries, 
is the happening of " something unexpectedly, unintention- 
ally." Suppose, instead of containing typhoid germs, as in 
the present case, the water that the employees of the assured 
consumed had contained some of the most virulent poisons, 
would anyone contend that the injuries resulting therefrom 
could not be properly held to have been accidentally inflicted? 
We think not, and yet, in our opinion, there is no substantial 
distinction between the case supposed and the case at bar. INDEX 
PAGE 
Acid, fatty, liberation of by steam 474 
Acid waters, attack lead 460 
Aerated layer, extent of in lake.., 309 
Aeration and agitation, efficiency of 223 
Aeration and iron removal 210 
Aeration by air-lift 210 
Aeration, filtration, and copper sulphate for algae control 335 
Aeration of water 208 
Aeration, shakes out bottom odors 325 
Aerator at Kensico for New York City 224 
Aerator at Springfield, Mass . 209 
Age at which typhoid commonly occurs , 95 
Age, relation of typhoid death rate to 95 
Agitation and aeration, efficiency of 223 
Agitation, not favorable to growth of diatoms. 318 
Air, analysis of . . 244 
Air binding of sand bed 179 
Air, comparison of country and city 246 
Air, dissolved, ratio of oxygen to nitrogen in 272 
Air lift, aeration by. 210 
Air lift for Asbury Park water 438 
Air of Paris, bacteria in 246 
Air of Paris sewers .. 246 
Air, ratio of oxygen to nitrogen in dissolved 474 
Air, soot in, effect of 244 
Air wash, compressed air for 169 
Air wash for gravity filter 170 
Albany, cost of water purification at 146 
Albany, deaths due to typhoid in 96 
Albany, filter beds, section of 130 
Albany filters, area of 121 
Albany, monthly distribution of typhoid at.... 90 
Albany, plan of filters at , . 126 
Albany, size of clear-water basin •. 132 
Albany, sundry data for filter plant 144 
Albany, tax levied by typhoid fever 97 
Albany, typhoid follows submerging filters at 30 
Alcohol, formation of vinegar from 414 
Algae, food for, 311 
497 498 
INDEX 
PAGE 
Alg®, growth of controlled by aeration, filtration, and copper sulphate 335 
Algae, growth of in natural lakes 335 
Alg®, quantity of copper sulphate required to kill sundry 325 
Alg®, removal of odor by double filtration 335 
Alg®, varieties observed in certain cases of odor 325 
Alkalinity, required to decompose alum 488 
Aluminum hydroxide and coloring matter 175 
Alum, alkalinity required to decompose 488 
Alum, arsenic in 166 
Alum, danger to health from use of 166 
Alum, direct preparation of 166 
Alum, dry feed 165 
Alum, necessity for its use 160 
Alum required, influenced by turbidity 164 
Alum, specifications for 165 
Alum, theoretical dose required. . . . t 164 
America, average residual typhoid in 83 
Amount of wash water 169 
Amsterdam, Holland, water supply of 401 
Anab®na, growth of in Western reservoir 319 
Analysis of Hudson River water 262 
Analysis of Mississippi River water 263 
Analysis of water from Yellowstone Park 440 
Anchor ice 258 
Ancient Indian irrigating canals 3 
Ancient reservoirs in India 3 
Ancient reservoirs of the Singhalese 5 
Ancient Roman Castellum 9 
Anderson process 181 
Annular space, relation of to filter area 177 
Anthrax as a water-borne disease 108 
Anthrax conveyed by flies and dust 108 
Anthrax from water supply , 108 
Anthrax spores, killed by chlorine 108 
Aqueduct of Vienna 8 
Aqueduct to ancient Carthage. . 6 
Aqueducts, Roman 7 
Argo, case of the ship 12 
Army, latrine pit 364 
Army, sterilizing water for 220 
Arsenic in filter alum 166 
Arsenic in spring water 39° 
Artesian water, distant source of 429 
Artesian water, fluctuation of with tide 429 
Artesian water, supplying Jacksonville 442 
Artesian well, conditions favorable for 42$ 
Artesian well, Woonsocket, S. D 432 INDEX 
499 
PAGE 
Artesian wells, Charleston, S. C.. ' 429 
Artesian wells, conditions producing 419 
Artesian wells in Florida 427 
Artificial ice 259 
Artificial purification of water 116 
Asbury Park, deep-seated supply for 437 
Ashokan reservoir for New York City . 304 
Ashokan reservoir, cost of stripping site 332 
Asiatic cholera germ 54 
Assuan reservoir, Egypt 304 
Asterionella, excessive growth of at Brooklyn 319 
Asterionella, number of organisms required to produce odor 321 
Atlantic City, view of well basin at 313 
Attenuated and virulent virus, intensity of attacks by 86 
Auburn, N. Y., typhoid epidemic at 341 
Average value of human life 96 
Bacilli, attenuated, produce mild cases 104 
Bacilli, reduced in stored ice 253 
Bacillus Coli Communis, diagnostic value of 75 
Bacillus Coli Communis, effect of acids upon 68 
Bacillus Coli Communis in deep well-water caused by birds 76 
Bacillus Coli Communis in rock-drilled wells 447 
Bacillus Coli Communis more resistant than Bacillus Typhosus 74 
Bacillus Coli Communis more resistant than Cholera Spirillum 74 
. Bacillus Coli Communis, relation of to Bacillus Typhosus 63 
Bacillus Coli Communis, attenuation of by chlorine 193 
Bacillus Coli Communis, demands carbonic acid 219 
Bacillus Coli Communis, seasonal variation in 240 
Bacillus Prodigiosus experiment 230 
Bacillus Tuberculosis, action of chlorine upon 192 
Bacillus Typhosus, cultivated vs. natural 340 
Bacillus Typhosus, culture of, on agar 62 
Bacillus Typhosus, culture showing flagella 62 
Bacillus Typhosus, hibernation of 91 
Bacillus Typhosus, killed by storage in copper vessels 327 
Bacillus Typhosus, life history of, outside of the body 64 
Bacillus Typhosus, relation of to Bacillus Coli Communis 63 
Bacillus Typhosus, requires carbonic acid 219 
Bacillus Typhosus the cause of typhoid fever 60 
Bacillus Typhosus, ubiquity of 64 
Bacteria, action of chlorine upon 192 
Bacteria, alternate freezing and thawing fatal to 251 
Bacteria and spaces between sand grains, relative sizes of 153 
Bacteria, carbonated waters unfavorable to life of 68 
Bacteria, chlorine as dope for 193 
Bacteria, concentrated doses break down defense 103 500 
INDEX 
PAGE 
Bacteria eaters 311 
Bacteria, influence of turbidity upon deposition of 21 
Bacteria in deep waters 446 
Bacteria in Paris air 246 
Bacteria in river water less in summer 66 
Bacteria in streams, seasonal variation of. 239 
Bacteria in Thames water, monthly variation of * 240 
Bacteria not all killed by sunlight 66 
Bacteria, reduction of by storage 91, 339 
Bacteria, removal by sedimenting silt 341 
Bacteria, removal of, by precipitating mud 339 
Bacteria, toxic effect of sunlight upon 227 
Bacterial and chemical improvement not parallel 240 
Bacterial count, seasonal variation in 240 
Bacterial dose to produce observable effects 85 
Bacterial efficiency of settling tank 175 
Bacterial life, greater near shore of lake 306 
Baltimore, water used during great fire at , 456 
Bathing, consumption of water for 453 
Battles and rain-fall 273 
Berlin, typhoid mortality in 43 
Black rain 245 
Black waters, South American 10 
"Blaisdell" sand-washer 140 
" Bloody snow " .... 245 
Bog-waters of Ireland 10 
Boiler scale, character of 476 
Boiler scale from sea water 482 
Boiler scale prevention 483 
Boiler scale, slight film beneficial 480 
Boiler scale, wasteful of fuel.' 476 
Boiler tube, blocked by hard scale 476 
Boiler water, free acid objectionable 473 
Boiler water, objection to magnesium chloride in 474 
Boiler waters of Connecticut, classification of 482 
Boiler waters, opinions concerning 480 
Boiler waters, rating of 481 
Boilers, action of water upon 458 
Boilers, affected by water 473 
Boiling, advantages of 218 
Bore holes, method of making 421 
Bore holes of great depth, sundry 424 
Boston reservoir, stripping site for 330 
Boulders, trouble due to in drilling 424 
" Breathing wells " 433 
Brewing, water for 484 
Brooklyn method of washing slow sand bed 142 INDEX 
501 
PAGE 
Brooklyn, N. V., excessive growth of asterionella at 319 
Brown waters, ability to keep well n 
Bubbly creek water supply, purification of 112 
Burlington, Vt., diarrhoeal epidemic at 105 
Butler, Pa., typhoid epidemic at 106 
Calcium carbonate, protects lead from attack 462 
Calcium silicate, protects lead from attack 462 
Calcium sulphate, curve of solubility of 477 
Camp pollution, protection of reservoir against 365 
Camp, sanitary managing of 364 
Camp waste, incineration of 364 
Carbolic acid, odor and taste of in water 241 
Carbonated waters unfavorable to life of bacteria 68 
Carbon dioxide, action of upon iron 474 
Carbon dioxide at Lowell, removed by aeration 462 
Carbon dioxide attacks lead 460 
Carbon dioxide causes brilliancy of water 412 
Carbon dioxide, required by B. Coli and B. Typhosus 219 
Carbon dioxide tends to attack iron 469 
Carbon dioxide unfavorable to life of bacteria 68 
Care of water-sheds 361 
Carriers more dangerous in the country 102 
Carriers not limited to typhoid fever 102 
Carriers, typhoid 100 
Carriers, typhoid, agent in spread of disease 103 
Carthage, ancient aqueduct 6 
Carthage, ancient cisterns of 4 
Cave ice 258 
Charcoal, as to purifying effects of 217 
Charcoal filter, ancient use of 327 
Charleston reservoir, result of non-stripping 332 
Charleston, S. C., artesian wells of 429 
Chemical and bacterial improvement not parallel 240 
Chemical changes during stream flow 236 
Chemical precipitation at St. Louis 358 
Chemicals fatal to fish 271 
Chicago, betterment of water by intake tunnel 37 
Chicago drainage canal case 230 
Chicago drainage canal case, experiments with typhoid bacillus in 70 
Chicago drainage canal, data concerning 305 
Chicago, typhoid deaths in 234 
Chicago stock yards water purification litigation 112 
Chlorination beneficial at Erie, Pa 194 
Chlorination beneficial at Kansas City 194 
Chorination, cost of 194 
Chlorination, large scale experiments upon 193 502 
INDEX 
PAGE 
Chlorination, use of for sterilizing water for army 220 
Chlorine, action of, upon bacteria 192 
Chlorine as "dope" for bacteria 193 
Chlorine, attenuation of Bacillus Coli by 193 
Chlorine, dosing apparatus 191 
Chlorine in Troy snow • 260 
Chlorine, killing of Anthrax spores by 108 
Chlorine, liquid 190 
Chlorine, organic matter interferes with efficiency of 108 
Chlorine, taste, removal of 190 
Chlorine, variation of during snow storm 260 
Chlorine, variation of, in Troy rain. . 247 
Cholera a filth disease 59 
Cholera and washing of soiled clothing 23 
Cholera at Cuneo caused by laundry 25 
Cholera epidemic at Hamburg 49 
Cholera, epidemic at Messina 23 
Cholera in India 44 
Cholera, incubation period of 84 
Cholera Spirillum 54 
Cholera Spirillum, produces cholera, evidence that 57 
Cholera Spirillum, viability of 56 
China, water supply in 48 
Cincinnati, iron, sulphate, and lime process in use 162 
Cincinnati, value of sedimentation at 338 
Cisterns, filtering 249 
Cisterns, inspection and cleaning of 249 
Cisterns of ancient Carthage 4 
Cisterns, protection from mosquitoes 249 
Cisterns, suitable location for ,................... 249 
Cisterns with metal, objection to 248 
Cistern water, affected by cistern material 248 
Cistern water, analysis of, from dirty cistern 247 
Cities of the world, health status of leading 40 
City mains, influence of upon water 370 
Clark's process for softening water 478 
Claudian aqueduct, Rome •. Frontispiece 
Climate, affected by forest destruction 300 
Cleaning a reservoir while in commission 359 
Cleaning London filter beds 138 
Cleaning slow sand filters. 139 
Cleaning slow sand filters, frequency of 143 
Clear water basin, size of 132 
Clouds, formed of water drops 243 
Coagulant, amount used 163 
Coagulant imperative in mechanical filtration 160 
Coagulated material in settling tank 174 INDEX 
503 
PAGE 
Coagulation, pin-point 178 
Cochituate, temperature curves for Lake 314 
Cohoes, N. Y., data concerning plant at 163 
Coke trickier for iron removal 212 
Coke, use of for color removal 218 
Colorado River, material carried to sea by 263 
Color of water, effect of different lights upon 316 
Color of water, high in stagnant layer 315 
Color reduction in Wachusett reservoir 334 
Color removal, use of coke for 218 
Coloring matter and aluminum hydroxide 175 
Colorless water becoming popular 13 
Colorless waters, growing in favor 308 
Columbus, preparing reservoir site at 331 
Comma Bacillus 54 
Conduits, action of water upon 458 
Cone of influence, changes in •. . 389 
Cone of influence of well 380 
Connecticut River, rain-fall and river flow 292 
Connecticut, typhoid statistics for 40 
Consolidation of filter beds '... 165 
Consumption of water for bathing 453 
Consumption of water in metered cities 450 
Consumption of water in unmetered cities 449 
Consumption of water increased by plumbing extension 453 
Consumption of water increased by sewer introduction 453 
Copenhagen, water supply of 436 
Copper, attacked by soft water 467 
Copper, human organism easily tolerates 324 
Copper, in common articles of food 323 
Copper, toxicity greatly exaggerated 324 
Copper sulphate, applying to shallow bays 324 
Copper sulphate, cost of treatment with 326 
Copper sulphate, dose used at Troy 326 
Copper sulphate, fish killed by 324 
Copper sulphate process. 321 
Copper sulphate, quantity required to kill organisms 325 
Copper sulphate, safe limit for fish 326 
Copper sulphate, toxic to typhoid bacilli 326 
Copper vessels, ancient use of for disinfection 327 
Copper vessels, storage of water in, kills B. Typhosus 327 
Cost of chlorination 194 
Cost of cleaning St. Louis reservoirs .. . 358 
Cost of cleaning sundry slow sand filters 143 
Cost of construction of filter plants in general , 134 
Cost of constructing sundry slow sand beds 128 
Cost of maintenance of filters, general 133 504 
INDEX 
PAGE 
Cost of mechanical filter plant. . . ., 177 
Cost of municipal filtration, general average 146 
Cost of operating mechanical filter plant 178 
Cost of ozone treatment • 201 
Cost of Plymouth, Pa., typhoid epidemic 35 
Cost of preparing reservoir site at Columbus 331 
Cost of stripping Ashokan reservoir site 332 
Cost of stripping Wachusett reservoir site 330 
Cost of treatment with sulphate of copper 326 
Cost of water purification at Albany 146 
Counter-infection, spread of typhoid fever by 82 
Country and city air, comparison of 246 
Court of Appeals, decision as to reasonable contamination 363 
Covered filter at Ashland 129 
Covered reservoir, Pasadena 355 
Covers for slow sand filter beds 123 
Crenothrix 327 
Crenothrix, blocking of water pipe by 328 
Crenothrix, darkness favors its growth 329 
Crenothrix, iron required for growth , 328 
Crenothrix, objectionable to laundry interests 329 
Crenothrix, removal of, iron best guard against 329 
Crops, daily consumption of water by sundry 283 
Crowding, effect of upon health 94 
Crowding, influence of upon total death rate 94 
Cultivated vs. uncultivated typhoid bacillus 67 
Cuneo, cholera caused by laundry at 25 
• 
Dam, sub-surface 401 
Damages for serving unsafe water 98 
Danger to health from use of alum 166 
Danube River, seasonal variations of water of 262 
Death rate, general for London, improvement in 42 
Death rate, general for New York, improvement in 42 
Death rate, influence of crowding upon 94 
Deep-seated water 419 
Deep-seated water, B. Coli in 447 
Deep-seated water, characteristics of 439 
Deep-seated water, contamination of 445 
Deep-seated water, cost of drilling for 423 
Deep-seated water, drilling for / 422 
Deep-seated water, not inexhaustible 434 
Deep-seated water, not in every locality 438 
Deep-seated water of high color, case of 441 
Deep-seated supply for Asbury Park 437 
Deep-seated supply for Copenhagen 436 
Deep-seated water usually highly mineralized 439 INDEX 
505 
PAGE 
Deep wells, efficiency of 433 
Deep wells, freezing of 433 
Denver, Colo., skimming discs in settling tank at 345 
Depth, influence of upon character of water 308 
Depths of water on slow sand filters 135 
Detroit typhoid traced to Port Huron 229 
Dew-pounds 360 
Diamond drills, use of 421 
Diarrhoea, caused by vegetable overgrowth 308 
Diarrhoeal epidemics 105 
Diarrhoeal outbreaks, relation of to typhoid 104 
Diatoms, air essential to growth of 318 
Diatoms, flourish best on muddy bottoms 318 
Diatoms, growth of stimulated by light 318 
Diatoms, relation between growth of and circulation 317 
Diatoms, violent agitation not favorable to growth of 318 
Dilution as means of purification 232 
Disappearing streams 384 
Disease and drinking water 10 
Disease and filth, relation between 82 
Disease germs, antagonism of common bacteria to 69 
Disease germs destroyed by Plankton 70 
Disease, immunity to 84 
Disease may follow lowering of resisting power 82 
Disease, natural resistance to 104 
Disease, relation of to low-resisting power 82 
Disease, spread of through counter infection 82 
Disease, typhoid fever a preventable •. 94 
Disinfection of reservoir 219 
Disinfection of street mains 219 
Dismal Swamp water 11 
Dissolved gas, removal of in open heater 488 
Distilled water, wholesomeness of 114 
Distilled water plant at Gibraltar 208 
Distilled water, use on ships of the navy 115 
Does pure water pay? 94 
Domestic wells 385 
Drifting-sand filter................................................... 184 
Drilling, cost of 423 
Drilling for deep-seated water 422 
Drinking water and disease 10 
Drinking water and malaria ; 13 
Driven wells 391 
Driven wells at Lowell, Mass. 397 
Driven wells, delivery from 393 
Driven wells, dependent upon rain-fall 396 
Driven wells, French form 394 506 
INDEX 
PAGE 
Driven wells, mode of action of. , 391 
Driven-well plant, Brooklyn 392 
Droughts, historic 280 
Droughts in the Middle States, severe cases of 279 
Dry feed of alum 165 
Dunes, run-off of 403 
Dust, longevity of typhoid bacillus in 72 
Dust infection followed by typhoid epidemic 72 
Dutch filter beds, sections of 122 
Dyeing, water for 485 
Effective size of sand 154 
Efficiencies at sundry slow sand filter plants 148 
Efficiency of filters tested by use of bacteria............................. 151 
Efficiency of mechanical filters 178 
Efficiency of slow sand filters increase with use 148 
Efficiency to be looked for from slow sand filters 147 
Effluent regulator 136 
Elmira reservoir, result of stripping.. , 332 
Emergency intakes .. . 106 
Emergency intakes and typhoid fever 106 
Emergency purification methods 218 
Emergency water should be from safe source 107 
English filter bed system 118 
English cities, per capita consumption of water in 451 
Epidemics of typhoid and lowness of ground water . 93 
Epidemics of typhoid, ice-borne 253 
Erie, analysis of water of Lake 305 
Erie, Pa., chlorination at 194 
Erie, Pa., emergency intake at 106 
Erie, Pa., filter plant 168 
Erie, Pa., spot map of typhoid at 36 
Erie, Pa., typhoid epidemic at 105 
Erosion following forest destruction 300 
Evaporation, annual depth of 285 
Evaporation, depth of at Signal Service Stations 286 
Evaporation, effect of wind upon 281 
Evaporation, forest protection against 297 
Evaporation from forest ground 297 
Evaporation from water surface 281 
Evaporation from woodland soil 281 
Evaporation, maximum for the U. S 282 
Evaporation measurements 280 
Evaporation, rain-fall and flow of streams 273 
Evaporation, relation of to rain-fall, Mass 282 
Evaporation, surface tension of soil increased during 377 
Excess coagulation method. 179 507 
INDEX 
page 
Excess lime method of softening water 484 
Excess of area, slow sand filter i42 
Experience, argument of 99 
Experiments, demand for local 181 
Faith cures, recognized as curative 82 
Feed water, removal of dissolved gases from 479 
Filter, air wash for mechanical r7° 
Filter area, relation of annular space to i77 
Filter, automatic pressure • ; i59 
Filter bed consolidation 165 
Filter bed, effect of air binding of 179 
Filter bed, slow sand, section of 119 
Filter beds at Albany, section of ' 130 
Filter beds at Zurich 129 
Filter beds, composition of various 119 
Filter beds, covers for 123 
Filter beds, Dutch, sections of 122 
Filter beds, slow sand, area of sundry 121 
Filter bottom, Wheeler 172 
Filter, description of slow sand 118 
Filter, drifting-sand, view of 185 
Filter gallery 400 
Filter of charcoal, ancient use of 327 
Filter plant, Erie, Pa 168 
Filter plant, Norfolk, Va 167 
Filter, protects against concentrated bacterial doses 103 
Filter, Puech-Chabal 186 
Filter sand, weight of dirt upon 142 
Filtered waters, should be stored in the dark 312 
Filtering to waste, necessity for 152 
Filters, sponge, objection to 217 
Filters, which type to adopt 179 
Filth and disease, relation of 82 
Filtration employed during many ages 117 
Filtration, effective for odor removal 326 
Filtration lessens European typhoid death rate 149 
Filtration lessens typhoid death rate in New York 148 
Filtration, mechanical • • • 158 
Filtration, mechanical, amount of coagulant used 163 
Filtration, municipal, general average cost of 146 
Filtration not a straining process 153 
Filtration of sewage 222 
Filtration of surface water required by law 117 
Filtration, removal of odor by double 335 
Filtration, slow sand 118 
Fire at Baltimore, water used during 456 508 
INDEX 
PAGE 
Fire service, pressure required for 456 
Fire service, water required for 456 
Fires, relation of, to rain-fall 270 
Fish, dissolved oxygen used by 271 
Fish, effect of sundry sewage dilutions upon 271 
Fish, effects of sundry industrial wastes upon 270 
Fish, killed by copper in reservoir 324 
Fish, killing of by sewage or industrial waste 270 
Fish life and water pollution 271 
Fish life, oxygen required by major 272 
Fish, safe limit bf copper sulphate for 326 
Fish spawn, clogging of sand beds by 143 
Flies, protection of latrine pits from 365 
Floods, due to unusual causes 298 
Flora and fauna affected by sewage 241 
Florida, artesian wells 427 
Flow of streams, sundry causes of 290 
Fluorescein, use of for well testing 416 
Fly carrying typhoid fever 78 
Fly infection of food 79 
Fly distribution of typhoid bacteria 77 
Fly, life cycle of the house 77 
Fly typhoid epidemics, spot map 79 
Foaming, increased by use of permutit 486 
Foaming of boilers, cause of 473 
Food infected by flies 79 
Forest ground, evaporation from 297 
Forest protection against evaporation 297 
Forests act as governors of stream-flow 300 
Forests, destruction of, effects climate 300 
Forests, erosion following destruction of 300 
Forests, influence of upon flow of springs 297 
Forests, influence of upon water supply. . 295 
Forests, production of rain-fall by 299 
Forests, retard surface flow 284 
Forests, snow held longer in 297 
Frazil ice 258 
Freezing and thawing fatal to bacteria 251 
Freezing, as to purification by 225 
Freezing, beneficial effects of 253 
Freezing of sand filters, to be avoided 152 
Galvanized iron, attacked by water 463 
Gases, removal of dissolved from feed-water 479 
Gasoline, contamination of well-water by 415 
Gasoline, detection of by smell and taste 415 
Gatun Lake s 304 INDEX 
509 
PAGE 
German cities, per capita consumption of water in . 452 
Geyser, Old Faithful, water of . 440 
Gibraltar, data concerning . . ., . 351 
Gibraltar, distilling plant at 208 
Gibraltar, north front 350 
Gibraltar, storage of water at 349 
Gila River, variation in turbidity of ,263 
Goitre, relation of water to . 17 
Good, pure, wholesome water 109 
Gravity type of mechanical filter 161 
" Grease Balls " ....... 486 
Great Lakes, International Commission report 368 
Great Lakes ot North America - 5304 
Greece, sea mills of 447 
Grenelle Well, Paris 439 
Ground water 373 
Ground water, appearance deceitful 412 
Ground water, character of 387 
Ground water contaminated through soil cracks 409 
Ground water, contamination of by gasoline 415 
Ground water, contamination of from cesspool 412 
Ground water, determining rate of flow of 374 
Ground water, law as to ownership. . 416 
Ground water, manganese in 390 
Ground water, origin of 378 
Ground water of Western plains . .. .... . . . 382 
Ground water, serious contamination of 408 
Ground water, should be distributed quickly 337 
Ground water, typhoid fever and height of 91 
Ground waters, bad effect of open storage 312 
Ground waters, should be stored in the dark 312 
Hail, formation of.. 244 
Hamburg, cholera epidemic at 49 
Hamburg filters, area of 121 
Hardness of water and disease 15 
Hardness of water and typhoid, relation between 17 
Hardness of water, classification of 16 
Hardness of water, not injurious to health 17 
Hardness of water, relation of to health , 15 
Hard water forms lime soaps 484 
Hard water, kind of rocks supplying 444 
Hard waters, protection of lead by 460 
Health, as to effect of copper upon 467 
Health Board, rules of for water-sheds ............. 361 
Health, effect of crowding upon 94 
Health, effect of meter system upon 455 510 
INDEX 
' PAGE 
Health not impaired by hardness of water 17 
Health, relation of to hardness of water 15 
Health, relation of turbidity to 20 
Health status of world's, leading cities 40 
Heater, removal of dissolved gas in open 488 
Heligoland 378 
Highway pollution, protection of water-shed against 367 
Hippocrates, value of pure water 1 
Hospital drainage, protection of water-shed against 367 
Hourly consumption of water 136 
Hudson River valley epidemic of typhoid 27 
Hudson River water, analysis of.... 262 
Human beings relatively resistant to anthrax 108 
Huron, analysis of water of Lake 305 
Hydrochloric acid in river water . 266 
Ice, anchor 258 
Ice, artificial 259 
Ice, artificial, impurities in center of cake 252 
Ice as an article of food 250 
Ice, bacilli reduced in stored 253 
Ice-borne epidemics 253 
Ice cake, thickens downward 252 
Ice, examination of New York. . 256 
Ice field, flooding of 253 
Ice, formation in stand-pipe 357 
Ice formation, water concentration due to 252 
Ice, frazil 258 
Ice, from ice caves 257 
Ice, from Swiss glaciers 257 
Ice, law of Massachusetts concerning impure 250 
Ice, law of New York concerning impure 250 
Ice, needle 258 
Ice, rain, and snow 243 
Ice, rate of formation 257 
Ice removal from London filter beds. .. . 124 
Ice, safer in summer than winter 253 
Ice, strength of 257 
Ice, transparent, and snow 255 
Ice upon slow sand filters 123 
Illinois River, map of 231 
Immunity and tolerance 84 
Immunity gradually acquired 104 
Immunity, partial against disease invasion 103 
Immunity to disease 84 
Increased viscosity and loss of head 179 
Incubation period of cholera 84 INDEX 
511 
PAGE 
Incubation period of diarrhoea 104 
Incubation period of typhoid fever 84 
India, water supply and cholera 44 
Industrial waste, effect of upon fish 270 
Industrial waste pollution of rivers 267 
Industrial waters, character of 484 
Infection, spread of disease through counter 82 
Infiltration gallery 398 
Inherited tolerance of disease germs 84 
Insanitary surroundings, relation of to typhoid infection 81 
Intakes abandoned should be cut off 107 
Intakes, emergency 106 
International Joint Commission report 368 
Irish bog-waters 10 
Iron, action of carbon dioxide upon 474 
Iron, action of oxygen upon . 474 
Iron, amount required to grow crenothrix 328 
Iron, ancient monuments of 472 
Iron, as to presence of in water 467 
Iron, conditions tending to attack of 469 
Iron, objectionable in industrial waters 485 
Iron, quickly attacked by salt in water 471 
Iron removal at Asbury Park 437 
Iron removal at Copenhagen 437 
Iron removal by aeration 210 
Iron removal by use of air-lift , 210 
Iron removal, coke trickier for 212 
Iron, removal of best guard against crenothrix 329 
Iron, rusting of a complex process 471 
Iron stains, produced by crenothrix 329 
Iron sulphate, use of, because of cheapness 161 
Iron sulphate and lime process at Cincinnati 162 
Iron sulphate and lime process at New Orleans 162 
Iron sulphate and lime used in mechanical filters 160 
Iron, tuberculation of water-main 468 
Iron water pipe, life of 473 
Irrigating canals, ancient Indian 3 
Irrigation by sewage 222 
Isolated typhoid difficult to explain 83 
Jacksonville, Fla., artesian supply of » 442 
Kensico aerator for New York City 224 
Kerosene, detection of by taste and odor 416 
Lake Drummond water 11 
Lake Moeris, dimensions of . 3 512 
INDEX 
PAGE 
Lake, stream water flowing over surface of 344 
Lake, stream water hugging bottom of 344 
Lakes, growth of algae in 335 
Lascaris, teaching upon origin of disease 3 
Latrine pit, army form . / 364 
Latrine pits, protection of from flies 365 
Latrines, proper location for 365 
Laundry at Cuneo, causes cholera 25 
Laundry interests, objection of, to crenothrix 329 
Laundry, public, pollution of stream by 24 
Laundry use, water for . . 484 
Law as to ownership of ground water 416 
Laws requiring filtration of surface water 117 
Lead, action of water upon 458 
Lead attacked by acid waters 460 
Lead, calcium carbonate protects against attack upon 462 
Lead, carbonic acid attacks 460 
Lead cisterns, protection of 461 
Lead a cumulative poison 465 
Lead, moorland waters likely to attack 459 
Lead, new vs. old attacked by water 461 
Lead, organic matter serving to prevent action upon 459 
Lead poisoning and water supply 459 
Lead poisoning, symptoms of 458 
Lead, quantity in city supplies of Massachusetts 459 
Lead, quantity required to condemn a water 458 
Lead, soft waters attack 460 
Lead, solution and erosion of 460 
Libavius, discussion of potable water.. : 1 
Life, average value of human 96 
Light, influence of upon typhoid bacillus 65 
Light, stimulates growth of diatoms 318 
Light, ultra violet for water purification 202 
Lights of different colors, effect of upon colored water 316 
Lime soaps, formed by use of hard water 484 
Lime-softened water fatal to typhoid group 312 
Lime, softening water by excess of 484 
Lime to increase alkalinity 488 
Liquid chlorine 190 
London, amount of sewage of 22 
London filter bed, area of 121 
London filter bed, section of 119 
London filter beds, cleaning of 138 
London filters, statistics of r 128 
London, general death rate improvement in 42 
London Metropolitan Water Board, researches of 340 
London snow 260 INDEX 
513 
PAGE 
London water, volume of pumped from the ground 435 
Long Island, slope of water-table of 380 
Loss of head, slow sand filter 120 
Lowell, Mass., driven-well plant at 397 
Lowell, removal of carbon dioxide by aeration 462 
Lubricating oils, corrosion by 474 
Magnesium chloride, decomposition of by ammonium chloride 475 
Magnesium chloride, objectionable in boiler water 474 
Maidstone typhoid epidemic 408 
Malaria and drinking water 13 
Malaria due to mosquitoes 13 
Manganese in ground water 390 
Manganese, removal of by permutit •. 486 
Manganese, variety of crenothrix 329 
Matanzas sea spring 420 
Mechanical analysis of sand 156 
Mechanical filter plant, cost of construction 177 
Mechanical filter plant, cost of operating 178 
Mechanical filters, cost of construction 134 
Mechanical filters, cost of maintenance 133 
Mechanical filters, efficiency of 178 
Mechanical filters, open or gravity type 161 
Mechanical filters, rate of filtration 160 
Mechanical filters, washing 169 
Mechanical filtration 158 
Mechanical filtration, amount of coagulant used. 163 
Mechanical filtration, iron sulphate and lime as coagulants. 160 
Mechanical filtration, some form of coagulant imperative 160 
Median age in U. S., increase of 44 
Merrimac River, typhoid infection carried by 228 
Messina, epidemic of cholera . ... . .. •. 23 
Metallic solvency, conditions tending to ' 466 
Metals, action of water upon . . . . . 458 
Meter system, effect of upon health 455 
Michigan, analysis of water of Lake 305 
Milk, a source of typhoid fever -.. 77 
Milk epidemic, spot map of typhoid fever 77 
Milk, typhoid bacillus develops in . . 61 
Mine water, sulphuric acid in ... 474 
Mine water unfavorable to life of typhoid bacillus 68 
Mississippi River, composition of silt of 265 
Mississippi River water, analysis of.. . 263 
Mosquitoes cause malaria 13 
Mosquitoes, protection of cisterns from 249 
Mougiotia, occurrence of, upon London sand beds 123 
Mud, bacteria removed by precipitating 339 514 
INDEX 
PAGE 
Naples, reservoirs at 347 
Natural purification of water 221 
Natural resistance to disease 104 
Naval vessels, use of distilled water by 115 
Needle ice , 258 
Negroes susceptible to typhoid fever 60 
New construction, pollution of water by 108 
New Harrington water, contaminated through soil cracks 411 
New Haven, typhoid epidemic at 35 
New Orleans, iron sulphate and lime process at 162 
New York, amount of sewage per day 22 
New York, Ashokan reservoir for 304 
New York, average typhoid death rates in 83 
New York City, great pollution of river by 267 
New York, general death rate improvement in 42 
New York, typhoid fever lessened in state of 83 
New York, typhoid rates for 77 
Niagara Falls, improvement resulting from passage of 223 
Nichols' sand washer 13g 
Nitrification established in soil 221 
Norfolk, Va., filter plant 167 
Nuisance defined in New York law 269 
Nuisance, not always of same definition 269 
Nuisance, prevention of sewage inflow becoming 272 
Odor, number of organisms found in instances of 322 
Odor of asterionella, number of organisms required to produce 321 
Odors, aeration shakes out bottom 325 
Odors caused by organisms 14 
Odors due to oil globules 14 
Odors occurring in water 13 
Odors, removed by filtration. . 326 
Odors which occur in water 13. 
Oil globules, causing odors 14 
Oils, lubricating, corrosion by 474 
Oils in organisms cause odor 14 
Old Faithful geyser, water of 440 
Orleans, France, large springs supplying 421 
Owasco Lake, typhoid carried by water of 341 
Oxygen, action of upon iron . 474 
Oxygen, amount of in sea and fresh water 272 
Oxygen dissolved, influence of on typhoid bacillus 68 
Oxygen, dissolved, tends to attack iron 469 
Oxygen, dissolved, used by fish 271 
Oxygen, exhaustion of in stagnant layer 309 
Oxygen, putrefactive reaction from absence of 238 
Oxygen ratio of to nitrogen in dissolved air 272 INDEX 
515 
PAGE 
Oxygen required by major fish life 272 
Oysters, hibernation of 80 
Oysters, typhoid fever from 80 
Ozone as water purifier 195 
Ozone concentration 200 
Ozone, dose of 201 
Ozone, De Frise system 197 
Ozone, effect of on water 200 
Ozone, mixing with water 198 
Ozone sterilizing tower 196 
Ozone, the Martinikenfeld plant 199 
Ozone treatment, cost of 2or 
Ozone, treatment, efficiency of 202 
Panama, Gatun Lake 304 
Paper-making, water for » 485 
Paratyphoid fever 107 
Pasadena, covered reservoir 355 
Peaty water, effect upon health 10 
Permanently hard water, softening of 479 
Per capita consumption of water in ancient Rome 452 
Per capita consumption of water in English cities 451 
Per capita consumption of water in German cities 452 
Per capita consumption of water in metered cities 450 
Per capita consumption of water in unmetered cities 449 
Per capita consumption of water, suitable allowance 453 
Per capita consumption of water unreasonably large 452 
Per capita daily supply 449 
Permutit, increases foaming 486 
Permutit, removal of manganese by 486 
Permutit, softening water by 485 
Petroleum, taste of in water ' 242 
Phagocytosis, theory of 85 
Philadelphia filters, area of 121 
Philadelphia, value of sedimentation at 339 
Physical exhaustion, relation of to typhoid infection 80 
Pipes, action of water upon 458 
Pin-point coagulation 178 
Plague caused by bite of the rat flea 6r 
Plankton 308 
Plankton, as destroyers of disease germs 70 
Plankton, copper sulphate process to check 321 
Plankton, not all objectionable 311 
Plankton, number required to produce taste ' 320 
Plant food removed by Mississippi River 265 
Plant requirements in water 283 
Pliny, discussion of potable water 1 516 
PAGE 
Plymouth, Pa^, cost of typhoid epidemic 35 
Plymouth, Pa., typhoid epidemic at 32 
Plumbing extension increases water consumption 453 
Poisoning by lead, symptons of.. . . 458 
Poisoning by zinc chloride 466 
Poisoning by zinc sulphide 464 
Poisonously pure water 114 
Pollution of rivers by sewage and waste 267 
Pollution of stream, rights and duties regarding . . . . , 490 
Pollution of water often greater in winter 302 
Port Huron typhoid affects Detroit 229 
Portsmouth, Ohio, filter plant 171 
Potomac River, rain-fall and river flow 293 
Power of resistance, influence of environment upon 107 
Power of resistance, lowering of 104 
Pressure filter. 159 
Pressure required for fire service '. . . 456 
Protection of water-sheds 300 
Providence, R. I., minimum charge for water , 455 
Public purposes, water for 454 
Public, rights and duties of 492 
Purification by dilution 232 
Purification of water artificially 116 
Purification of water, natural . 221 
Purification of water, time an element in. 235 
Putrefactive reaction from absence of oxygen 238 
Quantity of per capita daily supply 449 
Railroad bed, protection of against pollution 367 
Railroad pollution, protection against 366 
Rain, abnormal 245 
Rain, at Troy, variation in chlorine of 247 
Rain, black 245 
Rain, red 245 
Rain, ice, and snow. 243 
Rain water attacks lead 461 
Rain-fall and battles 273 
Rain-fall and river flow for Connecticut River 292 
Rain-fall and river flow for Potomac River 293 
Rain-fall and run-off for sundry river basins 293 
Rain-fall and snow of the U. S 274 
Rain-fall and typhoid fever in Tees Valley 26 
Rain-fall, average for North-Temperate Zone 273 
Rain-fall, evaporation, and flow of streams 273 
Rain-fall, exceptionally heavy 273 
Rain-fall, greatest in tropics and near sea 273 
INDEX INDEX 
517 
PAGE 
Rain-fall, increases with elevation 273 
Rain-fall, monthly order of for Massachusetts . . 295 
Rain-fall, production of by forests 299 
Rain-fall, relation of evaporation to 282 
Rain-fall, relation of great fires to 274 
Rain-fall, volume of one-inch depth of. 273 
Rain-making by use of explosives ........... 273 
Rate of filtration, mechanical 160 
Rate of filtration, slow sand filters 135 
Rate of filtration, various measures of 137 
Rates of filtration at sundry slow sand filter plants 137 
Raw waters, fit and unfit, for purification 112 
Reasonable contamination, decision by N. Y. Court of Appeals 363 
Red water, character of . 486 
Reservoir, Ashokan, for New York City 304 
Reservoir, Ashokan, cost of stripping 332 
Reservoir, Assuan, Egypt . 304 
Reservoir, Boston Water Works, per cent of oxygen in 310 
Reservoir, board covers for, laying of 357 
Reservoir bottom, old better than new 335 
Reservoir at Charleston, result of non-stripping 332 
Reservoir, cleaning while in commission 359 
Reservoir, Columbus, Ohio, preparing site 331 
Reservoir, covered, Pasadena 355 
Reservoir disinfection 219 
Reservoir, Elmira, result of stripping 332 
Reservoir, Naples 347 
Reservoir, protection against camp pollution : 365 
Reservoir, protection of by policing water-shed 363 
Reservoir, Quebec 304 
Reservoir, removal of grass from 360 
Reservoir site, should be cleared if not stripped 335 
Reservoir sites, stripping of 329 
Reservoir sites, treatment of deposits of soft material 336 
Reservoir storage, advantages and disadvantages of 308 
Reservoir storage, bleaching effect of prolonged 334 
Reservoir, Wachusett, color reduction in 334 
Reservoir, Wachusett, result of stripping 333 
Reservoir, Wachusett, stripping site for ? 330 
Reservoirs, ancient Indian 3 
Reservoirs, clear water, size of 132 
Reservoirs, correcting shallow flowage in. 336 
Reservoirs, odor of sulphuretted hydrogen in 337 
Reservoirs, sedimentation, size of 131 
Residual typhoid fever .... c 83 
Resistance to bacterial invasion 104 
Resistance, influence of environment upon power of 107 518 
INDEX 
PAGE 
Resistance, lowering of, followed by disease 104 
Resisting power, lowering of and disease production 82 
Rights and duties enforced by statute 492 
Rights and duties of municipal corporations 491 
Rights and duties of the public 492 
Rights and duties regarding stream pollution 490 
Riparian owners, rights and duties of 490 
Rio Vanagre, acid in water of 266 
River and stream water 262 
River, clearing of by rurtning out 225 
River flow, material carried to sea by 263 
River water, bacteria fewer in summer 66 
River water, influenced by tidal action 267 
River water typhoid commonly mild ' 86 
River water, typhoid infection carried by 228 
River waters, unusual materials in 266 
Rivers as drains 400 
Rivers as drains, proper use of 116 
Rivers, considered as drains 268 
Rivers, contamination during flood 401 
Rivers, diminish in volume 400 
Rivers, pollution of by sewage and waste 267 
Rock temperature in deep wells 425 
Rocks, difficult to drill 422 
Roman aqueducts 7 
Roman Castellum, ancient 9 
Rome, per capita consumption of water in ancient 452 
Rome, water supply of ancient 7 
Roof, cause of water impurity 248 
Run-off and rain-fall for sundry river basins 293 
Run-off, in general for United States 292 
Run-off, map of for United States 295 
Run-off, quantity per square mile 290 
Rusting of iron a complex process 471 
Safe and suitable water . in 
Salt in water, increases action upon iron ~ 471 
Sand, character of 154 
Sand dunes, Holland ... 404 
Sand dunes, water from 401 
Sand, effective size of 154 
Sand from sundry filters, data concerning 157 
Sand, general specification for 157 
Sand grains, sizes of bacteria and spaces between 153 
Sand layer, thickness of 120 
Sand, mechanical analysis of 156 
Sand, uniformity coefficient of 155 INDEX 
519 
PAGE 
Sands, velocity of water in 374 
Sand voids require careful filling 152 
Sand-washer, Blaisdell's 140 
Sand-washer, Nichols' 139 
Sands and soils, water-holding powers of 377 
Sanitary inspection for water-shed 363 
Sanitary managing of camp 364 
Saratoga, deep waters at 419 
Schmutzdecke, composition of 118 
Scouring velocities 346 
Sea mills of Greece....-- 447 
Seasonal variations of water of Danube 262 
Sea spring off Matanzas Inlet 420 
Sea springs 380 
Sea springs off Holland 420 
Sea water, boiler scale from 482 
Sea waters, analyses of 489 
Secondary typhoid fever 83 
Sedimentation as purifying process 224 
Sedimentation, benefits of 337 
Sedimentation, percentage of matter removed in twenty-four hours 346 
Sedimentation, slower in winter 337 
Sedimentation, theory of clearing water by 346 
Sedimentation, value of at Cincinnati 338 
Sedimentation, value of at Philadelphia 339 
Sedimentation-basins, size 131 
Self-purification in Illinois and Michigan canals 237 
Self-purification of streams . 228 
Septic tank, longevity of typhoid bacillus in 72 
Settling tank, bacterial efficiency of 175 
Settling tank, coagulated material in 174 
Settling tanks with skimming discs. 345 
Sewage, amount of London 22 
Sewage, amount of per capita per day 22 
Sewage bacteria, influence of upon typhoid bacillus 71 
Sewage, broad irrigation 222 
Sewage dilutions, effect of upon fish 271 
Sewage disposal, relation of typhoid infection to..... 81 
Sewage filtration, should be intermittent 413 
Sewage, flora and fauna affected by 241 
Sewage, great volume of added to certain rivers 267 
Sewage material should breed no nuisance 269 
Sewage of New York, amount of per day 22 
Sewage or industrial waste, killing of fish by 270 
Sewage pollution of rivers 267 
Sewage, prevention of inflow becoming a nuisance 272 
Sewage, purification of by filtration 222 520 
INDEX 
PAGE 
Sewage, typhoid bacilli in filtered 233 
Sewage, volume of, to that of stream, relation of 269 
Sewer-air, influence of upon spread of typhoid fever 81 
Sewer introduction, increases water consumption 453 
Sewers, air of Paris 246 
Sewers, consumption of water before and after building 454 
Shell fish infection, remedy against danger of. 80 
Shell fish, typhoid fever from 80 
Sieves, standardized for sand analysis 155 
Silt, bacteria removal by sedimenting 34! 
Silt carried by Mississippi River 265 
Silt of Mississippi, composition of 265 
Silt of Mississippi, plant food in 265 
Singhalese, ancient tanks of 5 
"Sink hole " or " sunk well " 37g 
Slow sand filter, at Albany, sundry data for 144 
Slow sand filter covered, at Ashland 129 
Slow sand filter, description of 118 
Slow sand filter, excess of area for J42 
Slow sand filter, section of 119 
Slow sand filters, as to ice upon 123 
Slow sand filters, area of sundry 121 
Slow sand filters, Brooklyn method of washing 142 
Slow sand filters, clogged by algae 123 
Slow sand filters, cleaning of 139 
Slow sand filters, cleaning body of bed 142 
Slow sand filters clogged by fish spawn 143 
Slow sand filters clogged by vegetation 143 
Slow sand filters, cost of cleaning sundry 143 
Slow sand filters, cost of construction 134 
Slow sand filters, cost of constructing sundry 128 
Slow sand filters, cost of maintenance 133 
Slow sand filters, covers for.. 123 
Slow sand filters, depths of water upon 135 
Slow sand filters, efficiency of increases with use 148 
Slow sand filters, efficiency of, tested by use of bacteria 151 
Slow sand filters, efficiency to be looked for 147 
Slow sand filters, freezing of sand of 152 
Slow sand filters, frequency of cleaning 143 
Slow sand filters, ice removal from . . 124 
Slow sand filters, London, statistics of 128 
Slow sand filters, rate of filtration 135 
Slow sand filtration 118 
Snow, absorption of impurities from soil by , . 261 
Snow and rain-fall of the U.S „ . . 274 
Snow as' source of water supply , ... 25'9 
" Snow, bloody " .. , 245 INDEX 
521 
PAGE 
Snow, chlorine in Troy 260 
Snow, city and country 260 
Snow, formation of 243 
Snow, held longer in forests 297 
Snow, influence upon spring water 261 
Snow, London 260 
Snow, rain, and ice 243 
Snow storm, variation of chlorine during 260 
Soda ash to increase alkalinity 488 
Soft water, classification of 16 
Soft water, soda ash or lime added to 177 
Soft waters attack lead 460 
Softening of permanently hard water..' 479 
Softening water by Clark's process 478 
Softening water by permutit 485 
Soil bacteria, destructive effect on typhoid bacillus 74 
Soil, circulation of water in 373 
Soil, contaminated by privy vaults 408 
Soil cracks, ground water contaminated through 409 
Soil cracks, New Harrington water contaminated through 411 
Soil filtration, should be intermittent 413 
Soil, great purifying power of 413 
Soil layer, depth of, removed in stripping 330 
Soil, longevity of typhoid bacillus in 72 
Soil, nitrification established in 221 
Soil, rate of water-flow through 373 
Soil, surface tension increased during evaporation 377 
Soil, total surface of particles of 377 
Soils, physical properties of 373 
Soils, voids in 373 
Soils, water-holding powers of various 376 
Soldier, water needed daily by 456 
Solid matter in rain water 244 
Solution channel 379 
Soot in the air, amount of 245 
Soot in the air, effect of 244 
Southampton, softening water at 479 
Spanish War, fly infection of food during 79 
Specifications for filter alum 165 
Spirillum Cholera Asiatics 54 
Sponge filters, objection to 217 
Spores of anthrax very resistant 108 
Spot map of cholera epidemic at Hamburg 52 
Spot map of fly typhoid epidemic 79 
Spot map of milk typhoid epidemic 77 
Spot map of typhoid epidemic at Erie 36 
Springs at sea 380 522 
INDEX 
PAGE 
Springs, conditions forming deep 41g 
Springs, formation of 385 
Springs, influence of forests upon flow of 297 
Springs, poor showing of recently developed 390 
Spring water, analysis of 387 
Spring water, arsenic in 390 
Spring water, influence of snow upon 261 
Spring water, sulphuric acid in 390 
Spring water, zinc-bearing 390 
Stagnation, advantages and disadvantages 307 
Stagnation, not in itself objectionable 309 
Stagnant layer, formation of 309 
Stand-pipe, ice formation in 357 
Standardized sieves for sand analysis 155 
Sterilizing action of sunlight 65 
Sterilizing water for army 220 
Still water, purifies quicker than running water 307 
St. Louis reservoirs, cost of cleaning 358 
St. Louis, typhoid deaths in 234 
Storage, bleaching effect of prolonged 334 
Storage, ground water injured by open 337 
Storage of water at Gibraltar 349 
Storage, reduction of bacteria by 91, 339 
Storage, tantamount to sterilization 340 
Stored water. 304 
Strainers, type of, for wells 396 
Stream flow, case of lessening of on rainy days 290 
Stream flow, chemical changes during 236 
Stream flow, forests act as governors of 300 
Stream flow, sundry causes of 290 
Stream pollution, caused by watering stock 302 
Stream pollution, judged by local standards 268 
Stream water flowing over lake surface 344 
Stream water hugging bottom of lake 344 
Streams, as to purification of by sunlight 227 
Streams disappearing 384 
Streams, rain-fall, evaporation and flow of 273 
Streams, self-purification of 228 
Streams, self-purification of not certain .• 230 
Streams, sulphuric acid in 474 
Streams, surface of United States removed yearly by 266 
Streams, underground 380 
Street mains, disinfection of 219 
Stripping of reservoir sites 329 
Subsidence in rotated and in quiet water 176 
Sub-surface dam 401 
Sulphuric acid in mine water 474 INDEX 
523 
PAGE 
Sulphuric acid in river water 266 
Sulphuric acid in spring water - 390 
Sulphuric acid in streams 474 
Sulphuric acid present in water 243 
Sulphuretted hydrogen, odor in reservoirs 337 
Sunk well or sink hole 379 
Sunlight, ancient use of, for disinfection 327 
Sunlight, as to purification of water by 227 
Sunlight, sterilizing action of 65 
Sunlight, toxic effect upon bacteria 227 
Superior, analysis of water of Lake 305 
Surface water, improved by open storage. ...» 312 
Susceptibility, apparent increase in • 104 
Susceptibility of visiting strangers to polluted water 104 
Swamp, Dismal 11 
Swimming pools, care of 195 
Swiss lakes, ancient wooden piles in 329 
Tannates in river water 266 
Tannery waste likely to carry anthrax 108 
Tanning, water for 485 
Taste and odor in water 14 
Taste, number of organisms required to produce 320 
Tees Valley, rain-fall and typhoid fever in 26 
Temperature curves for Lake Cochituate 314 
Temperature of water, influence of, upon typhoid bacteria 340 
Temperature, usually high in deep water 439 
Temperatures in deep wells 425 
Thames water, monthly variation of bacteria in 240 
Thermal death point of typhoid bacillus 66 
Tidal action, influence of upon river water 267 
Tide, influence of upon artesian water 429 
Tile under-drain 125 
Time, an element in water purification 235 
Time of river flow vs. distance, importance of 268 
Tolerance and immunity 84 
Tolerance inherited or acquired 84 
Transpiration through leaves of trees 296 
Trees as pumping engines 283 
Troy, cost of copper sulphate treatment at 326 
Tuberculation of water-main 468 
Tuberculation, scraper used for removal of 469 
Turbidity, influences deposition of bacteria 21 
Turbidity, influence of upon alum required 164 
Turbidity, relation of to health 20 
Turbidity, variation of in Gila River 263 
Typhoid bacilli, accumulation in river mud 232 524 
INDEX 
PAGE 
Typhoid bacilli, copper sulphate toxic to 326 
Typhoid bacilli from scouring of stream bottoms 233 
Typhoid bacilli in filtered sewage 233 
Typhoid bacilli, reduced in stored ice 253 
Typhoid bacillus, cultivated vs. uncultivated 67, 340 
Typhoid bacillus develops in milk 61 
Typhoid bacillus destroyed by soil bacteria 74 
Typhoid bacillus, effects of acids upon 68 
Typhoid bacillus, experiments with in Chicago drainage canal case 70 
Typhoid bacillus, hibernation of 91 
Typhoid bacillus, influence of light upon 65 
Typhoid bacillus in water, longevity of 67 
Typhoid bacillus less resistant than Bacillus Coli Communis 74 
Typhoid bacillus, life history of outside of the body 64 
Typhoid bacillus lives longer in mud than in water 68 
Typhoid bacillus lives longer in water containing dissolved oxygen 68 
Typhoid bacillus, longevity of, in cold water 91, 340 
Typhoid bacillus, longevity of, in contact with sewage bacteria . 71 
Typhoid bacillus, longevity of, in dust 72 
Typhoid bacillus, longevity of, in glass flasks.. . . ; , 70 
Typhoid bacillus, longevity of, in Mendota Lake water 71 
Typhoid bacillus, longevity of, in septic tank 72 
Typhoid bacillus, longevity of, in soil ", 72 
Typhoid bacillus, mine water unfavorable to life of 68 
Typhoid bacillus on growing vegetables. 73 
Typhoid bacillus, poisoning power of, decreases with time 86 
Typhoid bacillus probably a saprophyte 61 
Typhoid bacillus, thermal death point of 66 
Typhoid bacillus, tolerance of 103 
Typhoid carried by water of Owasco Lake 341 
Typhoid carrier, agent in spread of disease 103 
Typhoid carriers 100 
Typhoid carriers, bill to provide for 102 
Typhoid carriers more dangerous in the country IO2 
Typhoid carriers with no typhoid history 103 
Typhoid, contracted during vacation 107 
Typhoid death rate improved by changing water supply 41 
Typhoid death rate in New York lessened by filtration 148 
Typhoid death rates at different ages. , 95 
Typhoid death rates for urban and rural Vermont 76 
Typhoid death rates for urban and rural New York 76 
Typhoid death rates improved in Massachusetts 38 
Typhoid death rates in cities of U. S 39 
Typhoid death rates in Europe lessened by filtration 149 
Typhoid deaths in Chicago and St. Louis 234 
Typhoid differentiated from typhus fever 60 
Typhoid epidemic at Auburn, N. Y 341 INDEX 
525 
PAGE 
Typhoid epidemic at Butler, Pa... 106 
Tyhpoid epidemic at Detroit, traced to Port Huron 229 
Typhoid epidemic at Erie, Pa 105 
Typhoid epidemic at large boys' school 64 
Typhoid epidemic at Maidstone 408 
Typhoid epidemic at New Harrington 411 
Typhoid epidemic at New Haven 35 
Typhoid epidemic at Plymouth, Pa 32 
Typhoid epidemic at Windsor, Vt 66 
Typhoid epidemic due to dust infection 72 
Typhoid epidemic follows submerging of Albany filters 30 
Typhoid epidemic in Hudson River Valley 27 
Typhoid epidemics from river water commonly mild . 86 
Typhoid epidemics, ice-bome 253 
Typhoid epidemics preceded by diarrhoeal outbreaks 104 
Typhoid fever a country disease 76 
Typhoid fever, age at which it commonly occurs 95 
Typhoid fever, at Albany, deaths due to 96 
Typhoid fever at Albany, monthly distribution of 90 
Typhoid fever a human disease 61 
Typhoid fever and emergency intakes 106 
Typhoid fever and height of ground water 91 
Typhoid fever and lowness of water in wells 92 
Typhoid fever, an accident 493 
Typhoid fever an autumn disease 89 
Typhoid fever and hardness, relation between 17 
Typhoid fever and influx of storm waters 93 
Typhoid fever a preventable disease 94 
Typhoid fever and rain-fall in Tees Valley 26 
Typhoid fever a water-borne disease 60 
Typhoid fever caused by Bacillus Typhosus 60 
Typhoid fever, danger of unrecognized cases 82 
Typhoid fever distributed by flies. . 77 
Typhoid fever distributed by milk 77 
Typhoid fever especially fatal to negroes 60 
Typhoid fever, from deep-seated water 446 
Typhoid fever from well and stream water compared 87 
Typhoid fever from shell-fish 80 
Typhoid fever, incubation period of 66, 84 
Tyhpoid fever, influence of sewer-air upon spread of 81 
Typhoid fever, isolated cases difficult to explain 83 
Typhoid fever, lessened in State of New York 83 
Typhoid fever mortality rate 0 
Typhoid fever mortality in Berlin 43 
Typhoid fever, period of convalescence p6 
Typhoid fever preceded by diarrhoeal epidemics. 105 
Typhoid fever, secondary and residual 83 526 
INDEX 
PAGE 
Typhoid fever spot map at Erie, Pa 36 
Typhoid fever, spot map of milk epidemic 77 
Typhoid fever, spread of by counter-infection 82 
Typhoid fever, tax imposed by 96 
Typhoid fever, unrecognized cases dangerous 82 
Typhoid fever, "vacation" , Io? 
Typhoid infection carried by river 228 
Typhoid infection, relation of to form of sewage disposal 81 
Typhoid infection, relation of to insanitary surroundings 81 
Typhoid infection, relation of to physical exhaustion 80 
"Typhoid Mary" IOi 
Typhoid, relation of to character of water 235 
Typhoid statistics for Connecticut 40 
Typhus differentiated from typhoid fever 60 
Ultra violet light apparatus 203 
Ultra violet light, cost of sterilizing with 204 
Ultra violet light, efficient in clear and colorless water. 228 
Ultra violet light for water purification 202 
Uncultivated vs. cultivated typhoid bacillus 67 
Under-drains, perforated tile 125 
Underflow, description of 383 
Underflow of semi-arid region 328 
Underflow, rate of movement of 382 
Underflow, reinforcement of 381 
Underground streams , . 380 
Uniformity coefficient of sand 155 
United States, surface of, removed yearly by streams 266 
Vacation typhoid IO7 
Value of a human life 95 
Vegetables, typhoid bacillus on 73 
Velocities required to move sundry materials 346 
Velocity of ground water flow 375 
Velocity of water through sands . 374 
Vertical circulation in lake 309 
Viability of the Cholera Spirillum 56 
Vienna, aqueduct of 8 
Virulent and attenuated virus, intensity of attacks by 86 
Viscosity, increased, effect upon loss of head 17g 
Voids in sub-soils 373 
Volume of sewage to that of stream, relation of 269 
Wachusett reservoir, color reduction in 334 
Wachusett reservoir, result of stripping 333 
Wachusett reservoir, stripping site for 330 
Warren, Pa., diarrhoeal epidemic at 105 INDEX 
527 
PAGE 
Washington filters, area of 121 
Washing mechanical filters 169 
Washing of soiled clothing and cholera 23 
Washing, public, pollution of stream by 24 
Washing, thorough, not desirable 169 
Wash water, amount of 169 
Wash water, velocity of 169 
Waste of water 455 
Water-borne anthrax 108 
Water, damages for serving unsafe 98 
Water for plant requirements 283 
Water for public purposes 454 
Water, good, pure, and wholesome 109 
Water in soil, circulation of 373 
Water movement and wind movement 343 
Water need, daily, for soldier . 456 
Water not a cause of malaria 13 
Water, odors occurring in 13 
Water, quantity of required by sundry crops 283 
Water required for fire service 456 
Water, safe and suitable in 
Water should be safe from the faucet 83 
Water supply in China 48 
Water table, definition of 380 
Water table, lowering of 395 
Water table, relation of ground surface to 381 
Water, taste and odor in 14 
Water, total quantity of, in earth's crust 434 
Waters, black, South American 10 
Waters, brown, ability to keep well n 
Waters, fit and unfit, for purification 112 
Water-shed, as to construction work within 364 
Water-shed, protection of against highway pollution 367 
Water-shed, protection of against hospital drainage 367 
Water-shed, protection of against railroad pollution 366 
Water-shed, sanitary inspector for 363 
Water-sheds, care of 361 
Water-sheds, protection of 300 
Watering stock as cause of stream pollution 302 
Well and privy vault, distance between 407 
Well at Tangier, Morocco 388 
Well, cone of influence of 380 
Well, contamination of from cesspool 412 
Well, in city, always suspected 413 
Well strainers, type of 396 
Well, surrounded by privies 406 
Well water and stream typhoid compared 87 528 
INDEX 
. PAGE 
Well water, contamination of 403 
Wells, " breathing " 433 
Wells, domestic 385 
Wells, dug near coast ; 378 
Wells, flowing 431 
Wells, of great depth, sundry 424 
Wells on small islands 378 
Wells, over-forcing 395 
Wells, poor showing from new 390 
Wells, result of heavy pumping of 396 
Wells, self-sinking ........ 385 
Wells, sinking of by live steam 391 
Wells, spacing of 398 
Wells, temperature in deep 425 
Wells, testing of for pollution 416 
Wells, typhoid fever and lowness of water in 92 
Wheeler filter bottom , 172 
Wind movement and water movement 343 
Windsor, Vt., typhoid epidemic at . 66 
Woonsocket, S. D., artesian well at 432 
Yellowstone Park, analysis of water from 440 
Youghiogheny River, acid in water of 266 
Zinc, attacked by water 463 
Zinc-bearing spring water .' 390 
Zinc chloride, poisoning by 466 
Zinc in sundry public supplies 464 
Zinc in supply of Brisbane 465 
Zinc in water, effects due to 465 
Zinc in waters of cities of Massachusetts 464 
Zinc not accumulative poison 465 
Zinc sulphide, poisoning by ...... 464 
Zinc, used to prevent boiler scale 483 
Zooglcea jelly, formation of 120 
Zurich, filter beds at . 129