(Compliments of the Author.} 
MECHANICAL APPLIANCES IN TOWN 
SEWERAGE. [Reprinted from the Journal of the Franklin Institute, April, l886.J 
MECHANICAL APPLIANCES in TOWN SEWERAGE. 
Geo. E. Waring, Jr. 
[A Lecture delivered before the Franklin Institute, January 18, /886.~] 
The general subject of town sewerage has been much discussed 
during these later years, and the importance of the adoption of 
some efficient measure for the removal beyond the limits of a town, 
while still in a fresh condition, of all those manifold wastes of human 
life and industry which are capable of being transported in running 
water, is well known. All this part of the subject is far better 
understood now than it was twenty years ago, or even ten years 
ago. It is hardly necessary to add further argument to the stimu- 
lus in this direction, that is so frequently and so sadly furnished by 
the repeated occurrence of diseases which, formerly ascribed to the 
act of God, are now well known to be due to the negligence of man. 
Perhaps the short time at our disposal this evening may, with 
advantage, be directed to the consideration of the mechanical appli- 
ances of which works of sewerage consist. Until recently we 
were guided almost entirely by tradition, " rule of thumb," and 
" common sense." 
Let us suppose that at the outset, in a new settlement, a com- 
munity established itself on both banks of a brook which could be 
depended on for a supply of water, and for the immediate re- 
moval of surface drainage. As the community grew, it resorted 
to the better source of water to be found at greater or less depth 
under ground, and the brook become only a drain. When it got 
to be a nuisance, from its irregular form, and as it became important 
to make use of the land along its immediate banks, its sides were 
walled. For the convenience of traffic, roads crossing it were 
bridged. In time it was found necessary, as a matter of con- 
venience and to get rid of its unsightly appearance, to cover it 
over. Little by little neighboring houses discharged their wastes 
into it, generally under restrictions against the introduction of 
fcecal matter, based upon the erroneous notion, so long prevalent, 
that fcecal matter was the only seriously offensive waste discharged 
from houses. Often this covered brook, now become a covered 
common sewer, remained for generations in that condition. Indeed 2 
such drains still exist in our oldest and largest cities, notably, one 
in Broad Street, New York, in the heart of the Wall Street district, 
which is still the only outlet of the drainage of a considerable popu- 
lation, but concerning the condition of which we have only tradi- 
tional knowledge. 
Very often water courses existing under the condition above 
depicted are not merely walled and covered over; their place is 
taken by large mason-work conduits constituting a greater or less 
improvement on the ruder device. 
This is the origin, and this the type, of town sewerage as gene- 
rally known in this country and elsewhere. The sewers of Paris 
had a similar origin and perform a similar office. Some of the 
larger ones among them, the only ones that the public is allowed 
to inspect, are applied to the further use of furnishing channels for 
water pipes, electric wires, etc. They were formerly used for gas 
pipes as well, but explosions due to leakage of gas made it neces- 
sary to abandon this use. 
Pretty nearly the whole of London is sewered on a similar sys- 
tem, and, as London and Paris are established authorities and 
trusted examples of municipal work, the system has extended else- 
where. The considerable use of large pipes as a substitute for 
brick conduits in London, and subsequently in other English towns, 
has furnished an example which has been followed here and 
abroad. 
The sewer of antiquity, even of comparatively recent times and 
in by far the majority of cases to this day, is constructed in a more 
or less modified form on the theory that it must remove surface 
wash from streets, the outflow from industrial establishments and 
the wastes of houses. 
Nearly forty years ago the idea was advanced in England that 
there would be a sanitary and economical advantage in sepa- 
rating street water-an uncertain and constantly-changing quan- 
tity-from the more regular flow of industrial and domestic wastes. 
This was the origin of the separate system of sewerage. Though 
then, and still, called by this name, the separation in English work 
has been only partial. It has been decided by the courts there that 
a house-owner cannot be compelled to make two drains from his 
premises. He is, therefore, allowed to discharge the rain water falling 
on his yard and on his roofs into the public sewer. This has the 3 
advantage-more important there than here, because of the much 
greater frequency with which it rains-of affording a certain irregu- 
lar flushing to remove deposits from the sewers. It has the dis- 
advantage-much less there than here, because of the infrequency 
of torrential rains in England-that it requires the sewers to be 
constructed of an unduly large size as compared with what would 
be necessary for the immediate purpose of removing foul wastes. 
The English separate system, which is in many respects a great 
improvement over that formerly prevalent in that country, has been 
proved by long experience to be reasonably well adapted to the 
conditions by which its use is regulated. 
The argument in favor of this separate system is a strong one. 
Large storm water sewers, however well they may be flushed 
during storms, are very liable to the formation of deposits during 
dry weather, because the mere thread of a stream that they ordi- 
narily carry is not sufficient to prevent the retention of organic 
matters which, during their retention, undergo putrefaction and give 
off objectionable gases. The admission of street wastes increases 
very much the amount of deposit, and, while matters coming from 
the street are less offensive under putrefaction than are ordinary 
household wastes, they are often bulky and heavy, and constitute 
a nucleus for the retention and accumulation of foecal matter, and 
other putrescible substances which, without their presence, would 
have a better chance to be washed away in the flow. Owing to 
their much larger size, these storm water sewers are more costly 
than the smaller separate sewers. As the ventilation of sewers, as 
generally constructed, is extremely imperfect, they often contain a 
confined atmosphere which, on the sudden increase of the volume 
of water during storms, is forced past the protecting traps into the 
street and into houses. Viewed merely as a channel for the dis- 
charge of water which would do harm if not properly removed, 
the storm water sewer, as now made, leaves little to be desired. As 
a channel for the removal of foul matters, it is, as a rule, completely 
effective only during heavy storms. Perhaps the worst charge that 
can be brought against the system is the extreme offensiveness of 
the contained air of such sewers. Any one who has had experience 
as I have had in the inspection of works of this character, cannot 
fail to have observed the almost absolute uniformity with which 
their atmosphere is contaminated by the products of putrefaction. 4 
Another serious objection to storm water sewers has been 
developed by the modern requirement which is imperative, in so 
many cases, to make some other disposition of their flow than its 
simple delivery into a river or harbor, or into the sea. There are 
a few favored localities, but they are very few, where such disposal 
is final and unobjectionable. The necessity is already realized in 
many places, and is fast being realized in others, either to extend 
and improve the outlet by pumping, or to purify the effluent more 
or less completely by mechanical precipitation, or by agricultural 
irrigation. If sewerage is to be pumped, it is important as a mea- 
sure of economy to reduce the amount to be so handled as much 
as possible; to this end, it is desirable that there be excluded from 
it all water which may without disadvantage be disposed of by 
natural flow. If the more obviously offensive portions of the 
sewage are to be removed by mechanical precipitation, it is import- 
ant that the volume be kept small, and that it be as nearly regular 
as may be, day by day. If it is to be used as a fertilizer, or is to be 
purified by distribution over land through which it may be filtered, 
it is especially important not only that the volume be kept small 
and regular, but that particularly during rainy weather, when the 
land is least capable of receiving it, its volume be not inordinately 
increased by adding to it mere surface flow which might find 
another and relatively inoffensive outlet. 
It is quite usual now, where storm water sewers are used, to 
make provision for the disposal of the increased flow during storms 
by means of relief overflows. This would serve a very good pur- 
pose if the storm water would be obliging enough to flow on top 
of the foul sewage, escaping at the overflow in a reasonably pure 
condition. Unfortunately, the whole mass becomes thoroughly 
mixed ; the foul flow is greatly diluted, and the overflowing excess 
is greatly contaminated. This is the most serious objection to the 
relief-overflow system, but there is another which, though gene- 
rally lost sight of, is by no means unimportant. The overflow weir 
must be fixed at a certain height; unless it is so high that it will 
surely retain in the main sewer the maximum flow of an average 
day, which occupies ordinarily but a very short time in the fore- 
noon, some of this maximum flow will escape daily at the overflow 
outlet. If, to provide against this, the overflow weir is placed 
somewhat above any possible elevation of dry weather flow, then 5 
even a very light, continuous rain, will increase the daily discharge 
at the main outlet to very many times the volume due to house 
drainage alone. 
Where the English separate system is used, supposing the same 
restrictive requirements to obtain, a similar method of relief over- 
flows must necessarily be adopted, producing the same difficulties, 
though perhaps in somewhat less degree. 
Another serious disadvantage, which has been recognized only 
within a few years past, and which applies in even greater degree 
to the smaller English separate system than to the storm water 
system, is due to the enormous amount of air carried into a sewer 
by the rapid flow of roof water down the leader pipes. This flow 
attains great velocity and force. It never occupies anything like 
the full bore of the leader, except during very heavy storms, and 
at or near the top. In its descent it carries forward, with force, a 
great amount of air; this air, entering the sewer, becomes im- 
pounded and accumulates under such pressure as to force a final 
relief, often with great power. Instances have been cited where 
sewage driven back through a house drain, has spirted up through 
a kitchen sink with such force as to strike and besmear the ceiling; 
and where iron man-hole covers, luted somewhat securely into their 
position with dirt, have, on giving way under the increasing pres- 
sure, been blown high into the air. These effects and similar ones, 
even the perceptible blowing of air through the traps of fixtures 
in a house, are often due to the air carried in under accumulating 
pressure through roof leaders, and, to less degree, through street 
inlet basins. 
It is this effect of air-compression, added to the inadequacy 
of sizes, which may be sufficient in England, to cope with our 
violent rain bursts, which makes that system so unsuited for this 
country. 
If we assume that the systems above described are for any good 
reason necessary, and that relief from their disadvantages is to be 
gained by ingenious mechanical appliances, then we must confess 
that modern effort has been skilful and effective in giving relief. 
Some works with which I have myself had the good fortune to 
be connected would seem to indicate that relief is to be secured 
rather by avoidance and prevention than by remedy. 
Long impressed with the great disproportion between the sizes 6 
of sewers and the work that they had to do during at least ninety- 
five per cent, of the time, I had an opportunity to test the actual 
dry weather requirements, by gaugings made in a number of cities 
in different parts of the country. The result was conclusive, and was 
all in one direction. Among other records is that of a large sewer 
in Milwaukee, furnishing the only means of outlet for a population 
bf 3,1/7, having a very slight fall, and having the initial velocity 
of its flow arrested during the trial by a dam built across it. 
In this dam an ordinary six-inch sewer pipe was inserted. The 
greatest depth of flow during the forenoon of washing day-the 
greatest flow of the week-reached in this pipe a depth of only 
five and one-half inches. 
Based on these demonstrations, I laid out a system for the 
sewerage of Memphis, in which the size of all lateral sewers not 
more than 2,000 feet in length was restricted to six inches. There 
are now about forty miles of sewers on this system in Memphis, 
and of these more than eighty per cent, are only six inches in 
diameter. Unless accidentally obstructed, not one of these six- 
inch sewers has ever been known to run half-full; many of them, 
too, are laid on moderate grades. The success of this work has 
led to the construction of sewers of the same diameter in Omaha, 
Pullman, Ill., Kalamazoo, Little Rock, Leavenworth, Birmingham, 
A]a., Norfolk, Pittsfield, and Chelsea, Mass., Keene, N. H., and on 
a small scale in other places. A large experimental system on the 
same plan has been working constantly in one of the worst quar- 
ters of Paris for more than two years. 
If no other result had been accomplished by this work, it is 
much to have had it so universally demonstrated under such variety 
of conditions, that a six-inch sewer is fully adequate to the per- 
formance of the duty ordinarily imposed, so far as house drainage 
is concerned, on very much larger ones,- sewers costing more to 
construct1, more difficult to ventilate, and much more difficult to 
keep clean. 
In other words, considering a tube as the chief mechanical 
appliance for town sewerage, we have learned, so far as the removal 
of filth is concerned, that it will be greatly cheapened and 
improved by having its size reduced nearly to a minimum. 
We have learned by this part of the work that the simple func- 
tion of carrying a stream of sewage produced by a considerable 
population may be performed by a six-inch tube. 7 
If we consider some of the mechanical adjuncts of the Memphis 
work, we find that an equal improvement has been effected in the 
condition of the contained air of the sewer. A large tube, receiv- 
ing liquid and solid matters, has, at least during the upper part of 
its course, too shallow a current of liquid to carry all of the solids 
forward ; some of these are inevitably stranded by the way, other 
solids being gradually added to them, and all of the conditions 
favoring putrefaction. Putrefaction in a sewer is like putrefaction 
in a cesspool; it produces offensive and pernicious gases; it pro- 
bably develops the growth of dangerous micro-organisms, and it 
produces, as the sum of its results, that dangerous compound of 
stench and infection called " sewer gas." 
In the storm water sewer, as has already been hinted, the accu- 
mulation of deposits is favored by the introduction of rubbish from 
the streets. In the English separate sewers, the accumulation is 
considerable, because of the considerable radius necessarily given 
to the channel. In both cases, theoretically at least, each storm 
introduces a sufficient amount of water to wash these accumula- 
tions away; and in the best of such conditions during a storm the 
sewer is thoroughly cleansed. Unfortunately, no sooner does the 
rain cease than deposit and accumulation begin again, producing 
after a few days the former objectionable condition. So uniform 
has been this result that to most personsthe word " sewer" neces- 
sarily implies " sewer gas," and it is one of the most difficult tasks 
of the sanitarian to show that the supposed cause can exist with- 
out the assumed effect. 
As sewers were universally built before the Memphis work was 
done, the assumption was correct. There was introduced there, 
however, another condition which entirely altered the state of 
affairs. Having, because of the inadequacy of such small sewers 
to carry rain water, deprived them of even the occasional flush- 
ing that rain water would afford, it became necessary to adopt other 
means of cleansing. 
For this purpose there was placed at the upper end of every 
branch of the whole system a small brick reservoir, built under 
the street, with a capacity of about two hogsheads, fed by a trickling 
stream from the water supply, and provided with an automatic 
siphon by which, when the tank was filled and began to overflow, the 
whole volume was rapidly discharged into the head of the sewer, 8 
assuring it a thoroughly cleansing current throughout all of its 
upper portion, where the natural stream is never strong enough at 
any time during the day to remove all that is delivered into it. 
The source of foul air was thus very largely, but not with abso- 
lute completeness, removed; there remained still to be guarded 
against the smearing of the walls of the sewer which, without ven- 
tilation, would of itself be sufficient to produce a very foul condi- 
tion of the small amount of air that such sewers hold. The only 
means for obviating this difficulty seemed to be to secure a most 
copious ventilation, much more copious than would be possible 
with the large sewers of the combined system. 
The persistence of certain habits of thought has been curiously 
illustrated by those who have discussed the ventilation of the 
Memphis sewers on the basis of systems of ventilation applicable to 
large sewers. Even so distinguished a man as Professor Baumeis- 
ter has declared that, as air moves more freely through large 
channels than through small ones, therefore the ventilation of the 
Memphis sewers must necessarily be less effective than would be 
that of larger sewers with the same means of inlet and outlet, and 
with the same frictional forces acting upon it. He, and numerous 
others who have agreed with him, have disregarded the important 
consideration that, in order to make this reasoning correct, they 
should have given to the large sewer the same proportional means 
of inlet and outlet, and the same proportional effect of friction. For 
example: Take a six-inch sewer carrying a stream two inches deep, 
with a given velocity, having a length of 200 feet, and having eight 
occupied houses on each side of the way. Suppose these houses 
to be provided, as in Memphis, with untrapped four-inch house 
drains and soil-pipes opening into the sewer above the water-line 
and delivering above the roof with an open mouth. Under such 
conditions we have in this 200 feet of sewer about thirty-three 
cubic feet of air subject to the inward and outward currents due to 
sixteen four-inch pipes, having a combined sectional area of about 
200 square inches. The area of surface of a stream flowing two 
inches deep in a six-inch sewer 200 feet long is about ninety square 
feet. As the stream flows, it produces a frictional effect on the 
volume of air above it, which, in order to renew the whole atmos- 
phere of the sewer, has to impart its movement to only thirty-three 
cubic feet of air. 9 
Let us now suppose the sewer to be three feet in diameter and 
to carry the same volume of water. The flow, having much less 
hydraulic mean depth, will have less velocity, probably enough to 
counterbalance its increased area in contact with the air, so that it 
will be fair to ascribe to it the same influence in moving the con- 
tained volume, but that volume is now increased from thirty-three 
cubic feet to 1,414 cubic feet. As a much greater area of the wall 
of the sewer will be covered and laid dry with fluctuations in the 
flow, the amount of smearing will be greater and the amount of 
contamination will be in proportion. This will, in part if not 
entirely, make up for the greater dilution which the larger volume 
affords; so that if we consider only the effect of the stream we see 
that the purification of the contained air can by no means be so 
great. 
In the case of the six-inch sewer, the proportion between the 
square feet of ventilating area furnished by the sixteen house- 
drains and the cubic feet of contained air, was as one to twenty-five 
(about); in the case of the three-foot sewer it is as one to over 
i,OOO. Assuming, as we safely may, that the force applied in the 
case of the six-inch sewer is very much greater than is needed for 
the purification of the six-inch pipe, we see that purification 
is practically complete; and observation in the case of all 
six-inch sewers constructed on the Memphis system, of which 
I have had knowledge, shows, not a comparative but a complete 
absence of the peculiar odor of organic putrefaction. That is to 
say, in such sewers "sewer gas" does not exist, and the fear en- 
gendered by the present almost universal condition of town sewers 
becomes entirely unfounded. 
In addition to the effect of ventilation through house drains and 
of the friction of the stream ordinary flowing, a sudden greatly in- 
creased effect is produced by the discharge of the flush tanks. 
These send through the sewer such a volume of water, flowing at 
such high velocity, that practically all of the air in advance of the 
wave is driven forward with force and sent out through the house 
drains as they are successively reached, the partial vacuum follow 
ing the wave being supplied by a sudden indraft through those 
which have been passed. 
Incidentally to the discussion of the Memphis system, it is ap- 
propriate to speak of another of the mechanical appliances of 10 
sewerage known as the manhole. The manhole is a well, built 
from the sewer to the surface of the street and there properly 
covered, through which access is obtained to the channel for obser- 
vation and for cleansing. Guided, as we are all guided, by fixed 
traditions, I planned to construct manholes at regular short inter- 
vals over the whole work. When the estimates came to be made, 
it was found that the total cost of construction would be more than 
the community was able to pay; something had to be cut down. 
It seemed that the safest thing to dispense with would be the man- 
hole on the small lateral sewers. This effected a very great saving 
and, with the help of a restriction of the size of the main sewer 
to what was necessary for present purposes, the total cost was 
brought within the prescribed limits. 
Curiously, the chief criticism that has been made on the system 
here adopted related to the absence of manholes-which was not 
a part of the system at all. The subsequent modification of the 
work to which the abandoning of manholes led, was the introduc- 
tion of handholes at short intervals. It was found that when 
obstructions occurred, and they were very easily located by their 
effect on house drains, it was necessary to break through the top 
of the pipe to remove them. The openings thus made could not 
be satisfactorily closed. There was then introduced, in the further 
work, a pipe having an opening at its top and a moulded cap of 
the same material to fit it. The location of these caps being 
known, it was always easy to get access to the sewer within not 
more than fifty feet of an obstruction. Later, some of these open- 
ings were covered with caps having branches into which six-inch 
pipes were inserted, the standpipe being carried up to within about 
two feet of the surface of the ground. This greatly facilitated the 
work of inspection, the street being little disturbed by excavating 
to this slighter depth. These appliances have been effective and 
have now become a regularly adopted feature of the system 
wherever applied. 
Owing to obstructions due to special causes, there have been 
built in Memphis six manholes on six-inch sewers; on one lateral 
within a length of 700 feet, there are four manholes, and on another 
there are two manholes 300 feet apart; these are the only ones on 
about 180,000 feet of six-inch sewer, being an " average " of one to 
about five or six miles. So far as I have been able to learn, there 11 
is no disposition to build more. At. the same time, but for the 
cost of the work it would unquestionably be an improvement to 
build frequent manholes leading to but not wzto the pipe. Such 
small sewers are so easily obstructed that an opportunity should 
not be given to throw into them things which it may be desired to 
conceal. The sewer pipe should be exposed at the bottom of the 
manhole, but entrance to it should be through a handhole with a 
well secured cover. 
As a substitute for manholes, handholes and standpipes, a recent 
improvement is worth consideration. It is the invention of Burton 
R. Phillipson, of Dublin. He introduces at intervals a branch with 
a funnel-shaped junction with the sewer, so as to afford no occasion 
for floating objects to lodge, and he continues this with an inclined 
pipe nearly to the surface of the ground, covering it there with a 
special casting, taking the place of the ordinary manhole cover 
and easily opened for inspection of the sewer. As this pipe 
inclines toward the axis of the sewer and has a rounded connection 
with it, a stout telegraph wire, which is after all the best device 
for inspection or for cleansing, can be inserted, in the absence of 
angles, for a distance of several hundred feet. Incidentally to this 
arrangement, Mr. Phillipson has invented a very ingenious device 
for cleansing the sewer or damming back water for special flush- 
ing effect in case of need. He uses a soft india-rubber ball, or 
elongated bag, capable of being expanded to rather more than the 
full size of the sewer; to this an air pipe is connected; it is 
inserted into the sewer a little below the branch, and is then blown 
up until it makes a tight fit; the pressure of water accumulating 
behind it only makes it fit the tighter and in this way sewage may 
be dammed back to the surface of the street. Any partial obstruc- 
tion higher up the stream may then be driven from its attachment 
by the forcible working of a plunger in the vertical column of 
water, imparting the shock to the contents of the sewer above. 
When the proper flushing volume has accumulated, the opening 
of the air valve will speedily release the rubber plug which may 
be drawn out, allowing the whole volume to flow forward. 
Still following the traditions, the public, many of our boards of 
health, and not a few of our engineers, still insist that whatever 
else is done and whatever other safeguards are provided, the 
drainage works of the house must be absolutely shut off from any 12 
form of connection with the atmosphere of the sewer by a trap on 
the main drain. 
It would be out of place here to enter into a full discussion of 
the subject, but it seems pertinent to observe that this is another 
case where prevention is better than remedy; and that, disre- 
garding all other bearings of the case, if the universal ventilation 
of the sewer through house drains will secure the abolition, or the 
very great mitigation, of the sewer gas difficulty, it is better to 
accomplish this in this way than to allow the dreaded demon to 
exist, as he surely will in the absence of such ventilation, and then 
try to keep him out of the house by the interposition of a few 
quarts of water. 
It is also to be said that all communication between the house 
itself and the pipe through which its wastes are discharged must 
be shut off by traps at all fixtures, which shall be absolutely 
reliable under all circumstances. This being done, even very foul 
air coming from the sewer will pass out through the open soil- 
pipe, rather than pursue a tortuous and obstructed course through 
small branches into our rooms. 
A cardinal element of the system used at Memphis is the auto- 
matic flush-tank, already referred to, placed at the head of each 
branch sewer of the whole system. It is quite important that 
whatever device may be used for flushing the sewers should be 
arranged to work automatically and with certainty. It not only 
costs inordinately, in money and in water, to flush sewers by hand 
through their manholes, but the performance of an essential duty 
like this, the neglect of which, even for forty-eight hours, would 
establish a foul condition in the sewers, depending on personal 
attention, and on the direction and control of the average under- 
paid and overworked city official, cannot be relied on. 
Automatic flush-tanks are of various sorts and kinds. Perhaps 
the oldest and most familiar is the tumbler-tank, which becomes 
overloaded on its discharging side when nearly full, and then tips 
and spills, dropping back to its former position after it has been 
emptied. This tank is costly, for its joints and bearings must be 
rust-proof; its balance and counterpoise must be nicely adjusted, 
and, however carefully it may be made, it is pretty sure, sooner or 
later, to get stuck up or down, or more often at the half-way 
point, and to refuse to act. 13 
There have been many recent inventions of flush-tanks, some 
of them very ingenious, and some of them very good and practical. 
The only one that has come to my notice which seems to be abso- 
lutely unobjectionable is what is known as Rogers Field's tank» 
Mr. Field is a skilful and careful engineer, who has long made a 
specialty in England of works of sanitary drainage. The siphon 
by which this tank is discharged has been with him a plant of 
slow growth, and its ultimate perfection was the result of a long 
series of experiments. It has the very great advantage-one 
might almost say, considering the importance of having everything 
connected with a system of sewerage so adjusted that it will take 
care of itself, the indispensable advantage-that it is entirely 
without moving parts. It is practically as simple as a mill dam. 
Some minor modifications of form have been made in this country, 
as the result of experience and of observation, but the original 
principle remains precisely as Mr. Field devised it. Its peculiarity 
is confined entirely to the discharging limb, and it makes practi- 
cally no difference whether its form be a bend, or what is known 
as an annular siphon. Let us consider the annular form: 
A reservoir of a certain size is provided with a standpipe, pass- 
ing through its bottom and having its top at the level to which it 
is desired to fill the tank. This standpipe, for a tank of ordinary 
size, is four inches in diameter. Its lower end under the tank is 
not connected directly with the head of the sewer, but delivers 
into a basin, which, when filled to its overflow point, holds water 
just in contact with the lower end of the standpipe, sealing it, 
but not sealing it to any considerable depth. The upper end of 
the standpipe is provided with an adjutage, or funnel, from which 
the falling water drops clear of the sides of the pipe, and falls 
directly into the water in the lower chamber. As it falls it carries 
bubbles of air with it. Some of these bubbles rise again into the 
standpipe, but some are driven beyond its edge and rise into the 
chamber outside of the pipe. So long as this remains an 
uncovered standpipe, this process might go on forever without 
withdrawing more water from the tank than would naturally over 
flow at its top. 
Let us now take a dome, or a larger pipe, say eight inches in 
diameter, tightly closed at the top, and pass it down over the 
standpipe so as to leave a space between the two for the whole 14 
circumference. This space now becomes the receiving limb of a 
siphon, and, if siphonic action is once established, it will continue 
to run until the level of the water in the tank is carried down to 
the bottom of the outside pipe and air is taken in. There is noth- 
ing novel in the annular form of siphon. 
What takes place in Field's siphon is this: The top of the 
standpipe being sealed by the water contained between it and the 
outer pipe and the lower end being sealed by contact with water in 
the discharging chamber, we have, at the moment when the over- 
flow begins, a certain volume of confined air at natural atmospheric 
pressure. A small stream overflowing and dropping from the 
adjutage to the water below, carrying air bubbles down as described, 
and some of these bubbles escaping outside of the foot of the stand- 
pipe, there is a gradual removal of the air; the pressure is reduced, 
water rises a little into the lower end of the pipe and the rapidity 
of overflow is proportionately increased. The increased flow car- 
ries out more air, and after a few moments' operation-longer or 
shorter according to the volume of the stream added to the reser- 
voir-so much is removed, and the falling stream is so much in- 
creased that the remaining air is soon discharged, and the siphon 
works with a solid stream. 
In Memphis, the tanks have a discharging capacity of 120 gal- 
lons-about two hogsheads. If accurately adjusted, a stream 
barely large enough to fill the tank once in twenty-four hours is 
enough, after it is filled and begins to overflow, to bring the siphon 
into action within from five to ten minutes, and the entire volume 
is discharged, in some cases in as little as thirty-five seconds. 
It was found that while the siphon could be brought into action 
as described, something further was needed to make its breaking 
complete; that is, to get rid of all of the water contained in the 
siphon itself after the first checking of the flow by the first intake 
of air. This could best be done from below. It is accomplished by 
the operation of what is called a subsidiary siphon. This, in its sim- 
plest form, is a pipe carried from the discharging chamber over its 
elevated overflow point and down into the outlet. While the main 
volume is being discharged, a part of it passes through this pipe, 
which continues to act as a siphon after the flow through the main 
siphon has ceased. Its further supply of water comes from the out- 
let chamber at the foot of the standpipe, the level of whose water 15 
is thus reduced below the lower end of the standpipe which, 
becoming unsealed, loses its entire charge, the contents of the re- 
ceiving limb dropping back into the tank, and the contents of the 
discharging limb falling into the chamber and passing off through 
the subsidiary siphon. 
The water of the subsidiary siphon having been lowered an 
inch or more below the mouth of the standpipe, a new discharge 
of the tank cannot be set up until it has been refilled and the over- 
flow has continued sufficiently long to fill the outlet chamber again 
and thus reseal the lower end of the discharging limb. 
It has been found in practice that a small pipe adequate to the 
emptying of the discharging chamber is subject to stoppage from 
the accumulation of rubbish of one sort or another before its inlet. 
This difficulty was obviated by using a diaphragm in the main 
discharge which, while not materially reducing the capacity of the 
outlet, gave a subsidiary siphon of sufficiently large area to carry 
off foreign matters without choking. 
Then another difficulty arose. The outlet of the subsidiary 
siphon being so large, air was taken back through it and its action 
was stopped before the desired effect was accomplished. To cover 
this difficulty a block or dam is placed in the outlet in front of the 
subsidiary siphon, holding back the flow to such an extent as to 
prevent the admission of air until the work of emptying the outlet 
chamber is done. 
Still another difficulty was found in practice: As the standpipe 
or discharging limb was originally made, it was but little larger 
than the mouth of the adjutage or funnel. Blisters of rust accumu- 
lating on the inside of the pipe sometimes projected so far out as 
to catch the falling stream and lead it down into the outlet chamber 
along the side of the pipe without carrying out the air. To remedy 
this, the pipe was made considerably larger, increasing the space 
between the edge of the funnel and its outer wall, and then drawn 
in at the bottom so that the water should fall very near to its side 
and allow a larger proportion of the air bubbles to escape outside. 
These modifications and some improvements in construction not 
necessary to describe here having been made, the siphon is now 
absolutely and permanently effective. It is simple in construction 
and may be depended on to perform its offices year after year with 
absolute certainty. It may be, of course, that the ingenuity that 16 
is constantly applied to the subject will result in producing some 
other form of automatic device that will be even cheaper and better 
than this; but thus far nothing has been produced at all com- 
parable with it. 
There is a good field for invention, and much need for it, in the 
matter of outlet, especially for the drainage of low-lying districts 
and of town areas which, while they may have sufficient fall to low 
water mark, have their outflow obstructed twice a day by the rise 
of the tide. 
One of the most ingenious things that has been devised in this 
connection is what is known as Shone's ejector, somewhat used in 
England, and now contemplated for application at Sacramento. 
This is a sort of self-acting compressed air pump, controlled by 
automatic valves operated by floats. A tight reservoir of any con- 
venient size is built in the ground at the lowest point of the dis- 
trict it is to serve, and to this the sewage of the whole district is 
led. Its outlet communicates with a force main; its air vent is 
open and its outlet is closed. The sewage flows into it, and when 
it is filled, operates floats, which close the air vent and admit the 
pressure by which the contents are driven through the force main. 
This system of pumping is especially applicable to towns having 
numerous depressed points in no easy communication with each 
other, or for very large, wide areas where sufficient fall cannot be 
obtained without a depth of cutting that would be very costly in 
unstable soils. Its chief drawback relates to the loss of power in 
compressing air, which is serious; and to its moving valves. 
Generally it would be cheaper, maintenance considered, to lay 
one large, low main sewer, by tunnel if necessary, to carry the 
out-fall level back with little loss for fall, or to do the pumping all 
at one central station, applying the power directly to the pump 
rather than to use it for the compression of air. 
Perhaps, too, the same end may be accomplished by the direct 
application of power at as many different points as may be neces- 
sary, by means of gas engines. These are about being used for the 
pumping of the whole sewage flow of the borough of Stamford, 
Conn., and that experiment will give us the information necessary 
to determine whether or not this power may be economically 
applied at different points throughout a town. 17 
A case has recently occurred in the carrying out of a plan of 
my own, in which it was necessary to surmount an unexpected 
obstacle. 
In Norfolk, Va., the entire sewage of the city is discharged into 
a pumping well, in which the mouth of the suction pipe stands at 
a point ten and one-half feet below low water mark. The main 
sewer of a system covering about one-third of the town was to be 
discharged into this well, and in order to reach branches neces- 
sarily laid at about the level of high water in the most distant part 
of the town, a low grade for the main sewer had to be preserved. 
It was contemplated to lay for this a pipe sewer eighteen inches 
in diameter about fifteen feet below the surface for a consid- 
erable part of its length. It was soon found that the soil in which 
it would be necessary to work was an almost impassible quicksand, 
and there was reason to fear that in passing a very large Masonic 
temple its foundations would be disturbed by the running of the 
earth. The scheme had to be abandoned. The sewers for the 
upper part of the system had already been laid, with the aid of 
active pumping. In devising the means for relief, I had recourse 
to a plan I had recommended for the drainage of a town in Hol- 
land, but not there put in practice. A well was sunk at the end of 
the completed main, about 1,700 feet distant from the pumping 
well, and was carried down several feet below the level of the intake 
of the pumps, its total depth being about twenty-one feet. Into this, 
the main sewer delivered. From it, at a point six inches below the 
sewer inlet, there started a fourteen-inch iron pipe, which was car- 
ried up the well and thence along the street, but little below the 
surface, turning down into the pumping well and being finished 
with a return bend, which would always remain full of water and 
would serve as a permanent sealing trap. The overflow of this 
sealing trap is a few inches higher than the intake at the open 
well. This constitutes a siphon by which the sewage flowing into 
the upper well is carried upward about thirteen feet and then for- 
ward into the pump well. As sewage rises in the upper well, it 
flows out at the mouth of the siphon, the head required for its 
discharge being practically what is due to the theoretical flow 
through a fourteen-inch pipe 1,700 feet long of the discharge of an 
eighteen-inch pipe, which will rarely, if ever, run half full. This 
siphon has now been working for some seven months, and it is 18 
said by the city engineer to be an entire success. It was charged 
by a pump attached to the main steam pumps and connected with 
the apex by a two-inch iron pipe. This pump is operated for a 
few minutes every day to remove accumulated air. 
Had I not recently had occasion to read a paper in this city on 
the disposal of sewage, including the destruction or utilization of its 
wastes, I should have regarded that subject as constituting a better 
compliance with the invitation of your Professor of Chemistry. 
The proper applications of the mechanical appliances used for the 
removal of the waste of population, need, however, to be well 
understood even by students who give their attention chiefly to 
the chemical problem connected with disposal and utilization.