Sewer Gas: 
Chemical^ Physical 
and 
Bacterio logical Studies. 
o 
By A. C. ABBOTT, M. D. 
{front Transactions of the Congress of American 
Physicians and Surgeons, 1804} 
... CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES 
UPON AIR OVER DECOMPOSING SUBSTANCES, WITH 
SPECIAL REFERENCE TO THEIR APPLICA- 
TION TO THE AIR OF SEWERS. 
By A. C. Abbott, M.D., 
First Assistant, Laboratory of Hygiene, University of Pennsylvania. 
[From the Laboratory of Hygiene, University of Pennslyvania. ] CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES 
UPON AIR OVER DECOMPOSING SUBSTANCES, WITH 
SPECIAL REFERENCE TO THEIR APPLICATION TO 
THE AIR OF SEWERS. 
BY A. C. ABBOTT, M.D., 
JVrst Assistant, Laboratory of Hygiene, University of Pennsylvania. 
[From the, Laboratory of Hygiene, University of Pennsylvania.'] 
Mr. President, gentlemen of the Congress : the subject set for dis- 
cussion this afternoon, on which I have the honor to speak, is one that 
has attracted attention from the medical and lay world for a number 
of years, and opinions upon it are still at variance. When I accepted 
the invitation of the President of the American Climatological Asso- 
ciation to present a paper at this meeting based upon experimental 
studies relating to the sewer gas question, I little realized the mani- 
fold ramifications of the task I had consented to undertake, and if the 
results of my efforts to obtain more light upon this important subject 
are not conclusive, and appear rather as indications than actual demon- 
strations, I look to your indulgence, and trust you may realize that a 
part, at least, of the shortcoming is attributable to the nature of the 
problem that has been attacked. 
I shall not trespass upon your time with a recital of the opinions 
that prevail with regard to the influence of the air of sewers, drains, 
and cesspools in the production of disease ; that you will hear from 
the speaker who is to follow me in the thorough way so characteristic 
of his writing, but shall ask your attention to a consideration of the 
chemistry, bacteriology and physics of the air of these localities. 
A review of the literature can but leave the impression that many 
of the opinions advanced from time to time upon these aspects of the 
subject are manifestly the outgrowth of intuition, or are ideas not 
based upon exact methods of investigation. From such sources we 
might be led to believe that the air of sewers is a mysterious mixture 
of poisonous, gaseous products of decomposition, holding in suspen- 
sion innumerable bacteria capable of producing all manner of disease, 
and existing under a tension that enables it to press outward in all 
directions, and especially into houses that are in communication with CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
29 
the sewer. On the other hand, opinions that are based upon the study 
of this problem by precise, scientific methods demonstrate the inac- 
curacy of such views at almost every point that the subject has been 
approached. We are now tolerably well acquainted with the nature 
of the air of sewers, thanks to the efforts of a few who have endeavored 
to learn the truth on the subject, and we know that as ordinarily found 
it does not differ very conspicuously from the air we are accustomed 
to breathe. As a result of the numerous and careful studies upon the 
chemistry of such air in situ by deClambray,1 Lethby,3 Nichols,8 Mil- 
ler,4 Russell,6 Beetz' and others, we learn that it is ordinary atmos- 
phere with which are mixed certain gaseous products of decompo- 
sition ; and that the most conspicuous chemical difference between it 
and pure air is that it is usually a little poorer in oxygen, and richer 
in carbonic acid, ammonia and sulphuretted hydrogen than is outside 
air. It was demonstrated by the investigations of these observers 
that the proportion of oxygen in the air of all sewers examined by 
them was never found to be lower than 17.4 per cent, by volume, with 
one exception, and this occurred in a choked arm of the notorious 
sewer Amelot in Paris where the reduction of this gas reached to 13.8 
per cent, by volume. The same observers found that the proportion 
of carbonic acid in the air of these sewers fluctuated between 0.12 per 
cent, and 3.4 per cent, by volume, and that frequently the other 
gaseous products of decomposition were either not to be detected with 
accuracy, or they were present in very small amounts. 
The studies that have been made by Miquel,7 Petri,8 and Carnelley 
and Haldane9 upon the air of sewers with reference to its bacteriology 
are concordant in demonstrating its comparative poverty in bacteria, 
their number becoming worthy of consideration only under such con- 
ditions as splashing or active bubbling of the sewage, and even then 
rarely or never reaching the figures obtained from analyses of the air 
of the overlying streets or of inhabited houses. 
A number of efforts have been made to determine if there exist in 
sewer air volatile organic matters that could be held accountable for 
the diseases believed by some to be contracted through the respiration 
of such air, and while different observers report the finding of such 
organic matters, the amounts in which they were detected are so small 
that they can readily be dismissed from consideration. 
We have been led to believe, through some of the teaching on this 
subject, that the air of sewers exists under an active pressure, and 
tends to force its way into houses in connection with the sewer. 
Those who have studied the conditions as they are found in sewers 
have thus far failed to obtain any evidence whatever in corroboration 
of such an opinion. 30 
SEWER gas: 
Professor Chandler10 in his lectures before the Trades School Plumb- 
ing class expressed himself on this point as follows : " The common 
idea that gas in our sewers exerts a pressure to get out is a fallacy ; I 
have yet to find a case to prove it." He continues-" I some time ago 
had a lot of pressure gauges made which were distributed among the 
students with the instruction to apply them wherever they could, but 
there has as yet been no case of pressure reported." He had a vent 
taken from the street sewer to his laboratory, thinking that then he 
could get sewer gas whenever he wanted it by simply turning on the 
stop cock, but no pressure has yet been indicated. He concludes : 
" The sewers are not tight enough to stand any pressure from gases. 
They are not even impervious to water." 
The well known observations of Rozahegyi11 upon the sewers of 
Munich demonstrated that in 172 observations there was on 57 occa- 
sions no movement whatever of the air in the sewers ; that for 45 
times when he found the current of air flowing toward the houses it 
was in the reverse direction 70 times ; and the studies of Santo Crimp12 
upon the movements of the air of the sewers at Wimbledon demon- 
strated the existence of currents away from the houses on 273 days 
against 97 days in which it was toward them, and 88 days in which 
there was not sufficient movement in either direction to be detected 
by an ansemometer. 
These studies upon the chemical, bacteriological and physical side 
of the subject together with many observations upon the health of 
those constantly exposed to the emanations from sewage, and other 
matters in which decomposition is in progress, are certainly not of a 
nature to warrant the opinion that sewer air is the dangerous factor 
that some believe it to be. 
Among the results obtained by those who have studied the condi- 
tions found in sewers by exact analytical methods there does not appear 
a single conclusive demonstration that the air of sewers stands in 
causal relation to the diseases for which it has been held accountable. 
Are we justified in considering these observations as conclusive? 
Do the observations made in sewers, as we find them, indicate all that 
is possible to occur in these places ? 
Is it possible to decide if the unimportant nature of this air in situ, 
as determined by chemical analysis, results from inactivity of the de- 
composition in progress in the sewage itself, or, is due to constant and 
rapid dilution of the products of active decomposition by air from 
without ? 
Can we, through laboratory methods, offer any arguments that may 
be of service in formulating a reasonable view upon the bacteriological CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
31 
condition of this air, and its probable rdle in the spread and dissemi- 
nation of infectious diseases ? 
Is it possible to demonstrate experimentally on animals that the 
gases arising from sewage or from other decomposing substances have 
any direct effect upon the health of the animal, or its ability to resist 
infection ? » 
These represent some of the problems that we have endeavored to 
solve by experiments in the laboratory, problems that can only be sat- 
isfactorily answered when practically all surrounding conditions are 
under the control of the experimenter. 
No argument is required in support of the statement, that when one 
enters a sewer or smells of sewage the impression made upon the 
senses is very different from that obtained from such actively decom- 
posing substances as an infusion of meat, or other bodies rich in pro- 
teids, and if sewage is confined in a closed place there is often no 
evidence to the senses of active decomposition, and one cannot avoid 
the conclusion that the changes going on in sewage must frequently 
be relatively inactive, or else, different from those that accompany 
ordinary putrefaction. I have endeavored to satisfy myself as to the 
accuracy or falsity of this opinion by parallel experiments made with 
sewage, and with solutions known to undergo active decomposition. 
The results of these experiments, will, I think, suffice to demonstrate 
the relative inactivity of the decomposition frequently in progress in 
sewage as compared with matters more favorable to the process. 
They also indicate by their variability the impropriety of preconceived 
general notions on the subject. We find that with certain samples 
decomposition is so feeble that after as long as 14 days there will 
still remain as much as 17 per cent, by volume of available oxygen in 
the air of the flask and the amount of carbonic acid will have in- 
creased to only 1.42 per cent., whereas with other samples in which 
the sewage is mixed with slime scraped from the sides of the sewer, 
the gaseous exchange is much more conspicuous, in one case being 
quite as pronounced as that seen in meat infusions undergoing putre- 
faction under similar conditions (see Table I). These differences are 
hardly to be explained through differences in the character of the 
microorganisms, to the activity of which decomposition is due, for, 
if we remember the studies of Roscoe and Lunt13 upon the bacteria of 
sewage, they demonstrate that the majority of the different species 
isolated by them possess the property, when placed under conditions 
favorable to growth, of conspicuously reducing the amount of oxygen 
in the overlying air, and increasing coincidently the proportion of 
carbonic acid. The studies of Hesse14 have shown that this is a prop- 32 
SEWER GAS : 
erty not peculiar to putrefactive bacteria, but is one that is possessed 
by a number of other forms not normally concerned in putrefaction. 
Evidently then the limited gaseous exchange that we observe, at least 
in some samples of sewage, is owing to unfavorable conditions offered 
by the sewage itself. 
From another series of experiments additional evidence of the inac- 
tivity of the decomposition in sewage has been obtained. If one con- 
nects a mercurial monometer with a flask containing an actively de- 
composing body, such as an infusion of meat, one observes after a 
very few hours that a change in the level of the mercury has taken 
place. There is usually at first a slight depression of the mercury, 
due doubtless to the primary extraction of oxygen. This is quickly 
followed by an increase of tension, which, after a few days becomes 
conspicuous, often reaching in from 6 to 8 days a positive pressure of 
as much as 120 or 130 millimeters. I have tested two samples of sew- 
age by this means; one consisted only of the fluid sewage as it 
flowed from the sewer, while the other was a mixture of this fluid 
with slime scraped from the bottom of the sewer. In neither case was 
there an excursion of the mercury much beyond what could be 
accounted for by fluctuations in the temperature of the surrounding 
air. The results of these observations as compared with a similar ex- 
periment with putnfying meat infusion are represented graphically in 
chart I, and, as will be seen, give testimony in favor of the opinion of 
Professor Chandler already quoted. In the light of experiments 
of this kind, conducted under circumstances most favorable to the 
activities of the factors concerned in causing an increase of gas ten- 
sion, it is difficult to maintain that the air of so pervious a vessel as a 
sewer could acquire, as a result of decomposition in the sewage, suffi- 
cient pressure to enable it to force itself outward for any great dis- 
tance. 
During the course of this series of studies I have endeavored to 
detect the presence of the common products of decomposition other 
than carbonic acid in the air over sewage confined in sealed flasks, 
with irregular results. From many samples it was possible to obtain 
evidence of sulphuretted hydrogen, sometimes not in sufficient amount 
to detect other than qualitatively, while from other samples it was 
given off in relatively large amounts. In two samples none was de- 
tected, although the observation covered a period of six weeks in one 
and three months in the other case. Ammonia was practically always 
present, and varied conspicuously in amount. The air over several 
samples was tested for the presence of methyl-mercaptan, but I did 
not succeed in detecting it in any, though it was easily demonstrated CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
33 
in the air over actively decomposing meat infusions kept under similar 
circumstances. We have also endeavored to obtain a closer acquain- 
tance with the effect of ventilation upon decomposing bodies in much 
the same way that Erismann15 studied the gaseous products given off 
from decomposing faeces and urine. For sake of comparison these 
experiments were made upon decomposing meat infusion (50 grams 
meat to 1 litre of water) and upon sewage obtained from neighboring 
sewers. 
The experiments were arranged as follows : Triple neck Woulff bot- 
tles of four litres capacity were one-fourth filled with meat infusion or 
sewage as the case might be, thus giving a ratio between air space and 
material of 3:1. Into one neck of the flask was fixed, by means of a 
rubber stopper, a glass tube of about 5 mm. caliber, this reached to 
within 2 cm. of the surface of the fluid; into the middle neck was 
placed a thermometer and into the remaining orifice was fixed a short 
glass tube also of 5 mm. calibre. Both tubes were provided at their 
outer ends with hermetically closing stopcocks. All joints about the 
stoppers were sealed with paraflin. By this arrangement the air was 
caused to enter the flask by the longer, pass over the fluid and leave 
by the shorter tube. The air did not bubble through the fluid but 
passed over it. The entrance to the flask was guarded by an absorp- 
tion flask containing pumice stone saturated with a 50 per cent, solu- 
tion of potassium hydroxide, so that on entering the flask the air was 
free from carbonic acid. Between the exit tube of the flask and the 
aspirator, the absorption apparatus proper was arranged as follows : 
the air passed from the vessel containing the infusion or sewage first 
through two calcium chloride tubes, to dry it; it then passed through 
two U-formed absorption tubes containing pumice stone saturated with 
copper sulphate and dried, according to the directions of Fresenius for 
estimating sulphuretted hydrogen gravimetrically (see Fresenius' 
Quantitative Analyse, Vol. I, p. 505, also Zeit. fur analytische Cbe- 
mie, 10-75), it next passed through (during the first three or four 
days) Pettenkofers' tubes containing barium hydroxide solution ; these 
were, however, abandoned and 2 absorption tubes containing pumice 
stone saturated with 50 per cent, caustic potash solution were sub- 
stituted and the estimations made gravimetrically. To avoid error 
from loss of moisture these tubes were guarded at their exit by a 
layer of granular calcium chloride 4 cm. deep, separated from the 
pumice stone by a layer of asbestos. The air next passed to a gas 
meter that registered its volume in litres and thence to a Chapman 
water pump. 34 
SEWER GAS: 
The Pettenkofer tubes and barium hydroxide solution were dis- 
carded for the reason that, where large quantities of carbonic acid are 
given off, it is difficult to say when the solutions are saturated, and if a 
second and third tube be added to indicate this point we have no 
trustworthy means of determining if any, or what proportion, of the 
precipitated barium carbonate in the first tube is redissolved by the 
excess of carbonic acid passing through it. 
As the aspiration of air over the decomposing matters covered a 
number of hours between the estimations, usually about 24, and the 
temperature of the room was constantly changing, it was not thought 
advisable to reduce the air volumes to normal conditions from such 
data, so that in this respect there is a slight error in the results, an 
error that I hardly think could have been eliminated without having 
known the temperature of each litre, or small multiple thereof, that 
passed through the apparatus. 
In the accompanying tables (II and III) will be seen the data ob- 
tained from continuous observations made in this way upon air 
passed over decomposing meat infusion (Table II) and upon air passed 
over sewage (Table III). In Charts II and III these results are repre- 
sented graphically. 
As a result we find the amount of carbonic acid given off from sew- 
age is constantly smaller than that obtained from the meat infusion, 
while at times the proportion of sulphuretted hydrogen from the sam- 
ples examined became suddenly comparatively high. Whether the large 
amount of sulphuretted hydrogen given off by the sewage is due to the 
presence in the samples of sulphur compounds more easily converted 
into sulphuretted hydrogen than those in the meat infusion ; whether 
in the case of the meat infusion it was held back through combina- 
tions that it may have formed with other non-volatile or less volatile 
substances, or whether the difference is the result of peculiar fermen- 
tative processes in the one case that were not in operation in the other, 
as the observations of Petri and Maasen16 and of Stagnitta-Balisteri" 
have shown to be possible, I cannot say, the difference is, however, 
striking, as reference to the charts will show. With the sewage, as 
with the meat infusion, a sudden but temporary fall of temperature 
was almost invariably followed by a conspicuous increase in the 
amount of sulphuretted hydrogen in the air, while the carbonic acid 
remained unaffected (see tablesand charts). This sudden diffusion of 
sulphuretted hydrogen from the fluid, in which it, in part, seemed to 
be loosely dissolved, into the overlying air is doubtless to be explained 
through the difference of specific heat of the two bodies that are in jux- 
taposition, the sudden fall in temperature favoring diffusion from the CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
35 
fluid of high, to the air of lower specific heat. Another point of 
interest that is brought out in these experiments is that gentle agita- 
tion of the fluids likewise increased the amount of sulphuretted hydro- 
gen in the air, but had little effect on the carbonic acid. This is 
particularly the case with the sewage, where there was always a 
tendency to the formation of very dense pellicles on the surface, 
which seemed to prevent, when intact, the escape of this gas from the 
fluid. The breaking of such pellicles was almost invariably followed 
by an increase of sulphuretted hydrogen in the overlying air (see 
Chart III). When the solutions were kept perfectly still, and there 
was but little fluctuation of temperature, steady aspiration of the air 
over them showed a pretty constant production of carbonic acid, 
while the amount of sulphuretted hydrogen would often be reduced 
to nothing. A summary of these observations shows us that air 
passed over such an actively decomposing body as an infusion of 
meat, when ventilation of the chamber containing it is conducted at 
such velocities that the entire air is changed from 20 to 60 times in 24 
hours, carried off as a mean of 23 observations on as many days, 
29.96 parts per 10,000, by volume, of carbonic acid, the highest rela- 
tive amount on any single day being 75.97 parts per 10,000 with the 
ventilation equivalent to renewal of the air 24 times in 24 hours, and 
the lowest being 15.1 parts per 10,000 with ventilation at the rate of 
19 renewals of the air in 24 hours. When ventilation was reduced to 
practically nothing, that is, when the total volume of air was renewed 
only from one to three times in 24 hours, the mean amount of car- 
bonic acid reached the enormous proportions of 682.7 parts per 10,000. 
Air similarly aspirated over sewage at velocities that insured complete 
renewal of the contents of the chamber from 12.5 to 80 times in 24 
hours gave as the means of all analyses in each group of experiments 
the following relative proportions : 
Group 1-From December 11th to 30th inclusive=8.61 parts per 
10,000. 
Group 2-From January 10th to 19th inclusive=15.4 parts per 
10,000. 
The highest relative amount found was at the beginning of group 
2, when with a renewal of the air 28.8 times in 24 hours, 43.87 parts 
per 10,000 were obtained, while the lowest amount found was in group 
1, when with a renewal of the air 56 times in 24 hours only 3.31 parts 
per 10,000 were present. Expressing in the same terms the extremes 
of the results of analyses made in sewers by the observers quoted, we 
find that carbonic acid was found to fluctuate between 12 and 340 
parts per 10,000, while often other bodies were present in hardly 36 
SEWER gas: 
appreciable amounts. Under all conditions, therefore, not only in 
closed chambers, but under continuous ventilation (though varying 
in its rate), the decomposition in the samples of sewage examined, as 
evidenced by carbonic acid, its most conspicuous product, was usually 
of much less intensity than the decomposition of substances more 
favorable to the processes. 
When it finally became evident that the supposed poisonous prop- 
erties of the air of sewers could not be attributed to the substances 
that investigation had shown to be present, in the amounts that 
they were seen to occur, attention turned to the probable existence of 
a substance or substances, not included in these results, that might 
possess disease-inducing peculiarities. By some these bodies are 
assumed to be of a volatile organic nature, possibly closely related 
to the substituted ammonias ; by others they are said to be carbo- 
ammoniacal in nature ; by still others to be albuminoid ; but in fact 
little definite is known as to their existence or of their nature. From 
time to time methods have been devised for their isolation with the 
result that substances were obtained that gave certain reactions com- 
mon to organic matters, that is, of decolorizing potassium permangan- 
ate, or of being oxidized by it to ammonia. A detailed discussion 
upon the relative merits and accuracy of the various methods that 
have been employed would be out of place here. It will suffice to say 
that in the air over decomposing matters, just as in the open air, the 
air of rooms, and also in water, there may and do exist substances 
capable of reducing potassium permanganate, but which are not of 
necessity strictly organic, or of such a subtle nature that they cannot 
be easily detected by ordinary methods of analysis. The simple state- 
ment, therefore, that something has been found that decolorizes solu- 
tions of potassium permanganate, means little or nothing as to the 
nature of the substance. If, however, a body or group of bodies are 
found which, when oxidized, result in compounds that we know can 
come only from materials containing elements characteristic of certain 
groups, we then have a clue, at least, as to their general character. 
The oxidation, for example, of organic substances to ammonia leaves 
no doubt that the oxidized bodies were of nitrogenous nature. Advan- 
tage has been taken of this and methods have been devised for 
absorbing from the air its volatile principles and determining what 
proportion of them contain nitrogen, both in the form of free 
ammonia, and of potential or albuminoid ammonia-that is, nitro- 
genous matter convertible into ammonia, which latter is the body 
assumed to possess poisonous peculiarities. Of the various devices 
that have been recommended for their estimation the one suggested CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
37 
by Professor Remsen18 was selected as open to least objection, in so 
far as the chemical processes are concerned. It consists in aspirating 
the air that is to be studied through glass tubes containing granular 
pumice stone saturated with pure distilled water; these serve as 
absorbers for free ammonia, ammoniacal basis, and presumably other 
gaseous nitrogenous matters. When the desired quantity of air has 
been drawn through such absorbers their contents are transferred to 
carefully cleansed retorts or distilling flasks containing a measured 
amount of pure distilled water ; the free and albuminoid ammonia 
of this mixture is now determined by the well known method of 
Wanklyn, Chapman and Smith. By deducting from the totals now 
found the amounts of free and albuminoid ammonia previously deter- 
mined for the water used, the balance represents the amounts of 
these bodies that were present in the volume of air aspirated through 
the absorber. 
There are many details to be observed in the performance of this 
method : the flasks and condensers must be absolutely clean and an 
estimate must not be undertaken until it is possible to obtain from 
the apparatus distillates that give no reaction for ammonia when 
tested with Nessler's reagent in nesslerizing tubes 31.5 cm. long and 
1 cm. internal diameter. (Recommended by Drown.19) The absorb- 
ing tubes 13 cm. long and about 8 to 10 mm. inside diam. must be 
cleansed in a hot solution of sulphuric acid and potassium bichromate, 
after which is it our custom to rinse in hot water and finally pass 
steam through them for at least two hours. The pumice stone, broken 
and carefully sifted until the average size of all granules is about that 
of a linseed or a little larger, is heated in a covered platinum dish 
until red hot and kept at this temperature (in our experiments), over 
night, at least. It is then filled into the absorbing tubes. All rubber 
tubing, rubber stoppers, glass connections, etc., were cleansed first in 
boiling water and finally by passing steam through them. 
The nesslerization of all distillates and standards did not take place 
until they had stood, carefully protected from the air by steamed 
corks, for 18 to 20 hours in order that they might all be of the same 
temperature (see Hazen & Clark20), and it is advisable, before begin- 
ning the work, to make up a sufficient quantity of Nessler's reagent to 
last throughout the entire series of experiments. 
The distilled water used for dissolving out the materials contained 
in the absorbers must be carefully prepared, and in each experiment an 
analysis of the amount of water used in that experiment must be 
made, the result representing the correction (subtractive) to be made 
from the total results of the experiment. 38 
SEWER GAS : 
For further particulars we must refer to Prof. Remsen's original 
paper. (See bibliography.) 
Though objections have been raised to this method because of 
the extraordinary care necessary to avoid errors, it has, nevertheless, 
appealed to us as the one most likely to give concordant results when 
properly manipulated. Like all other methods that have been devised 
for this purpose, however, it affords no accurate information upon the 
nature of the substances retained in the absorber beyond the fact that 
they are or are not nitrogenous and are or are not convertible into 
ammonia by the processes to which they are subjected. 
We have applied this method to the study of the air over decom- 
posing meat infusions and to that over sewage, and for purposes of 
comparison, to outside air and air of the laboratory. The results have 
not however agreed with those of other observers using other methods, 
or with those of Professor Remsen himself using the same method. 
Throughout, our results have been uniformly lower than we had 
anticipated they would be. Of 31 estimates made upon the outside 
air, the air of the laboratory, air from over decomposing meat infu- 
sions and the air from over sewage, we have failed to obtain evidence 
of the existence of more than a trace ef gaseous, nitrogenous, organic 
matters, other than ammonia, the amounts being constantly so small 
as to fall far below the permissible limits of experimental error. 
Of 8 analyses of the open air obtained from just outside the second 
story of our laboratory not one showed the presence of nitrogenous 
organic matter, convertible into ammonia by the method used, though 
from 35 to 355 litres of air were drawn through the absorber. The 
amount of free ammonia was so low as to fall easily within the cate- 
gory of a " trace." 
Of 5 determinations made upon the air of my own room in the 
laboratory negative results were obtained in all, though from 31 to 
207 litres of air were drawn through the absorber. Free ammonia 
was present in barely a trace, and in one analysis none was detected. 
Of 6 experiments in which air was conducted over sewage, confined 
in a closed chamber, evidence of the presence of volatile, nitrogenous, 
organic matters was obtained in 3, the highest amount expressed as 
albuminoid ammonia being 0'004 of 1 milligramme for 36T litres ; 
while in the remaining three experiments no trace of such bodies could 
be detected. 
Of 7 experiments made by allowing air to bubble through sewage 
before passing to the absorber, the presence of volatile, nitrogenous, 
organic matters was demonstrated in 3, the remaining 4 being nega- 
tive. The highest amount found in terms of albuminoid ammonia 
was 0'005 of 1 milligramme for 866 litres of air. CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
39 
Of 4 experiments in which from 178 to 337 litres of air were aspi- 
rated over decomposing meat infusion, evidence of organic matter was 
obtained in all, the highest amount being O'O172 of 1 milligramme of 
albuminoid ammonia for 337 litres; while in one experiment in which 
143'2 litres of air were caused to bubble through decomposing meat 
infusion before passing to the absorber only 0'0019 of 1 milligramme 
of albuminoid ammonia was obtained. 
These figures are low, far lower than those given for organic matter 
in the air by other observers, and yet I am convinced that they are 
still higher than they should be, the excess being due to errors during 
manipulation, some of which can hardly be avoided, while others 
seem explainable. 
I account for the discrepancy between these results and those of 
Professor Remsen in part through the fact that in making the distilla- 
tions he assumed that all free ammonia was given off in the first 100 
c.c. of the distillate that passed over, and that all ammonia that now 
appeared, after the addition of the oxidizing agent, was derived from 
organic matter of an albuminoid nature.* 
In my experiments with sewage and with decomposing meat infu- 
sions, I was surprised to notice that, in the distillation of the con- 
tents of the absorbers for free ammonia, bodies that reacted with 
Nessler's reagent similarly to ammonia were often coming off at the 
end of 100 c.c. of distillation in such quantities that the process was 
continued without the addition of the oxidizing agent. The result was 
that in these analyses, particularly, this body often persisted until as 
late as 350 c.c. of distillation. The addition now of the oxidizing 
agent (alkaline potassium permanganate) showed that whatever was 
left in the retort was either not convertible into ammonia, or if so, in 
very minute traces. 
Because of the volatile nature of ammonia, we were inclined to the 
opinion that this body must have been something that gave with 
Nessler's reagent a reaction the same as that of ammonia but which 
was not ammonia pure and simple, differing in its physical properties 
at the boiling temperature. A series of control experiments, however, 
demonstrated that contrary to opinion, ammonia, as derived from pure 
ammonium chloride in the presence of sodium carbonate at the tempera- 
ture of 100° C., does not all come off in the first 100 c.c. distilled over, 
but is being given off up to the end of 200 to 350 c.c. of distillation, 
according to the amount originally present. (See Table IV.) In the 
same series of control tests a few experiments were also made upon 
* National Board of Health Bulletin. Washington, D. C., Sept. 11, 1880, Vol. 2, No. 
11, page 518. 40 
SEWER gas: 
methylamine and trimethylamine, alone and mixed together, and 
mixed with ammonia, and as the results show, they behave in all essen- 
tial respects identical with ammonia under the conditions employed. 
It is manifest that unless this point be borne in mind a large propor- 
tion of the ammonia obtained after the addition of the oxidizing 
agent, if this be done at the end of 100 or 150 c.c. of distillation, is not 
of necessity ammonia derived from bodies of an albuminous nature 
but is ammonia, or substituted ammonias, that could have been ob- 
tained without oxidation. It seems, therefore, probable that some, at 
least, of the albuminoid ammonia found in Professor Remsen's work 
may have been free ammonia or its substitution products. 
An example from one of several of my experiments may serve to 
make this point a little clearer and to demonstrate the possible error if 
what has been said is not taken into consideration. On Jan. 8th, '94, 
143-2 litres of air were aspirated through a decomposing meat infu- 
sion and thence through the absorber. The evidence of ammonia in 
large quantities was so conspicuous that 750 c.c. of pure distilled 
water was mixed with the contents of the absorber, instead of 500 c.c., 
as was more commonly the practice. The distillation for free 
ammonia was made in fractions of 50 c.c. each, and up to and includ- 
ing the 7th fraction, that is at 350 c.c., there was still evidence of free 
ammonia being given off, the 7th distillate containing 0'002 mg. The 
8th fraction gave no reaction for ammonia. The test for albuminoid 
ammonia was then made, resulting, after corrections for water used, 
in the total of- 
1.540 milligrams free ammonia 
0.0017 " albuminoid ammonia 
for 143.2 litres of air tested. 
Had, however, the results been calculated upon the assumption that 
all free ammonia had been given off at the end of 100 c.c. of distilla- 
tion, it is manifest that all that followed would have gone to the 
credit of albuminoid ammonia and the apparent amount of organic 
matter in the air would have been far greater than the case warrants. 
In short, the results calculated in this way would have indicated the 
presence of 
1.341 milligrams free ammonia 
0.2007 " albuminoid ammonia 
in the 143.2 
liters of air analyzed, in other words, the figures indicative of the 
presence of organic matter as expressed by albuminoid ammonia 
would have been more than a hundred times too high; and when the 
results of all my experiments are compared with what they would 
have been had I calculated albuminoid ammonia from the end of CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
41 
100 c.c. of distillation I find that throughout they would have been 
from 40 to 100 or more, times greater than they are. 
In the same way I propose to explain what I mean by saying that 
my own results are too high. A few examples will suffice for this. 
On Jan. 4, 1894, two experiments were made with air that had bub- 
bled through sewage. In the first, 320 litres, and in the second, 866 
litres were tested. I was not aware at the time of the persistency of 
free ammonia in the boiling solution so that the estimation for albu- 
menoid ammonia begun at the end of 200 c.c. of distillation. On the 
following day the fractions were nesslerized with the following results: 
1st Analysis-1st fraction = 0.352 
2d " = 0.060 
3d " = 0.012 
4th " = 0.0025 
Free ammonia total = 0.4265. 
1st fraction = 0.0045 
2d " = 0.000 
3d " = 0.000 
Albumenoid ammonia = 0.0045. 
lid Analysis-1st fraction = 0.440 
2d " = 0.120 
3d " = 0.024 
4th " = 0.006 
Free ammonia total = 0.590. 
1st fraction = 0.0055 
2d " = 0.002 
3d " = 0.000 
Albumenoid ammonia = 0.0075. 
Water used in these analyses. Subtractive correction for preceding 
figures. 
1st fraction = 0.008 
2d " = 0.0015 
3d " = 0 000 
Free ammonia, total = 0.0095. 
1st fraction = 0.0025 
2d " = 0.000 
3d " = 0.000 
Albumenoid ammonia = 0.0025. 
It is, I think, probable from these figures, that a portion, at least, of 
the results given for albumenoid ammonia really represents free am- 
monia. Just how much it is impossible to say. When this became 
evident 284.5 litres of air in one and 387.7 litres in another experiment 
were aspirated through the same sewage. The absorbers were mixed 
with 750 c.c. of distilled water and eight fractions of 50 c.c. each were 
distilled off for free ammonia with the following results : 42 
SEWER GAS 
1st Experiment-1st fraction = 0.120 
2d " = 0.D36 
3d " = 0.014 
4th " = 0.004 
5th " = 0.0015 
6th " = 0.000 
7th " = 0.000 
8th " =0.000 
Total free ammonia = 0.1755. 
1st fraction = 0.0039 
2d " = 0.000 
3d " = 0.000 
Total albumenoid = 0.0039. 
2d Experiment-1st fraction = 0.180 
2d " = 0.056 
3d " = 0.024 
4th " = 0.010 
5th " = 0.0025 
6th " = 0.0015 
7th " = 0.000 
8th " = 0.000 
Total free ammonia = 0.2740. 
1st fraction = 0.0038 
2d " = 0.000 
3d " ' =0.000 
Total albumenoid ammonia = 0.0038. 
The 750 cc. of water used contained- 
Free ammonia = 0.0143 
Albumenoid ammonia = 0.0038 
Subtractive correction for the two 
preceding analyses. 
It is not only seen that free ammonia was coming off as late as the 
fifth fraction in the first and the sixth fraction in the second experi- 
ment, but also that if distillation is continued until it ceases to appear 
before the test for albumenoid ammonia is begun, the figures indicative 
of the existence of this body become very much diminished. 
A comparison of the results of these two experiments, calculated for 
the same volume of air in both cases will, I think, impress this point 
more clearly. 
In the first pair of determinations there was evidence of albumenoid 
ammonia for 1000 cu. meters of air in the following amounts :- 
1st Experiment = 6.25 milligrams per 1000 cu. meters. 
2d " = 5.78 " " 1000 
While in the second pair, made with air bubbling through the 
same sewage, but with the point already made borne in mind, we 
obtained albumenoid ammonia in the following amounts :- CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
43 
1st Experiment = 0.352 milligram per 1000 cu. meters. 
2d " = 0.000 " " 1000 
The means of the results of each pair standing in the relation to one 
another as 34:1, an error that we think worthy of consideration. 
Still another reason for believing that my results are too high has 
been brought out in several experiments in which a group of absorbers 
were used, i. e. in which the air passed through a series of two or 
three absorbers. In these experiments care was taken to prepare all 
absorbers in identically the same manner and it was not infrequent to 
find, though in exceedingly small amounts, that apparently more albu- 
menoid ammonia was obtained from the second or third absorber than 
from the first. An example-on Jan. 12, 1894, 203.8 litres of air were 
drawn over a decomposing meat infusion and through two absorbers. 
Analysis of these absorbers gave for the first 0.0007 mg. and for the 
second 0.0022 mg. albumenoid ammonia, more than three times as much 
as was found in the first. 
In consequence I cannot avoid the opinion that, though the method 
has been performed with the greatest care, the results obtained by 
me for albumenoid ammonia are more apparent than real and that the 
figures presented for the amounts of organic matter as indicated by 
the albumenoid ammonia test are in part, at least, due to unavoidable 
error during manipulation. The results of these analyses are given in 
Table V, column 7. 
In addition to this method I have endeavored to employ the one 
suggested by Carnelley and Mackey21 in which the quantity of organic 
matter is determined colorimetrically by its decolorizing action upon 
solutions of potassium permanganate, but the results were so unsatis- 
factory that it was abandoned after a few trials. The color of the 
standard solutions recommended by them is so intense that errors of 
considerable magnitude may often exist without being recognizable ; 
moreover, in the air in which we were interested so many reducing 
bodies, other than those of an albumenoid nature, were present that 
decolorization of the solution exposed to such air afforded evidence of 
but little value. 
From the studies that we have made, and we consider them a fair 
test, it must be concluded that we have no good evidence of the exist- 
ence of volatile, nitrogenous, organic matters in the air other than 
ammonia or its substitution products. A series of experiments are 
under way to determine if it is possible to differentiate between am- 
monia and its substituted compounds in the small amounts that they 
appear in the air ; and while the outlook is promising I hardly think 
the work sufficiently far advanced to justify positive statements as 44 
SEWER gas: 
yet. It is manifest, I think, that the observation on the air of sewers 
in situ as well as those here presented, demonstrate that, as com- 
pared with other substances, the decomposition that goes on in 
sewage and the consequent pollution of the overlying air is much less 
than we are warranted in assuming. It seems also plain that we are 
not justified in forming anything more than a very rough conception 
of what is in progress, or what can occur under the conditions 
afforded by sewers and their contents. By the very nature of things 
these conditions undergo the widest range of variation. The compo- 
sition of the sewage itself varies not only in different sewers, but in 
the same sewer at different times in the day ; the dilution of the 
gaseous products of decomposition and fermentation, through the man- 
ifold portals of ventilation and diffusion, changes constantly with 
fluctuation in temperature, and particularly with the force-and direc- 
tion of the wind ; the pollution and dilution of the air of the sewer 
through leakage into it of air from the surrounding soil is not the 
same for all geological formations through which the sewer may run, 
and it does not seem unlikely that some of the waste products that 
gain access to sewers, especially from various manufactories may play 
an important part in influencing the activity or character of the fer- 
mentations that are to occur. When viewed from these standpoints 
we must admit the impropriety of conceiving the air of sewers to be 
a specific compound, or one that is even likely to approach constancy 
in its composition. 
When it long ago became evident that the potency of sewer air for 
the production of disease could hardly be explained by its chemi- 
cal composition, attention turned to its biological characteristics, and 
it was assumed because sewage contained countless bacteria, and more 
or less frequently bacteria of a pathogenic nature, that the air over 
such sewage must also of necessity contain these organisms. Since 
we have learned that many of the diseases formerly believed to result 
from the respiration of the air of sewers are of a specific nature and 
depend for their existence upon the presence in the tissues of particu- 
lar forms of micro-organisms, the opinion has been advanced that the 
part played by sewer air is that of a carrier of these infective parti- 
cles from the sewage to the individual who has become infected. 
Have wre any evidence in support of this opinion ? None that is 
based upon reliable methods of investigation. Unfortunately we 
know little or nothing of the nature of the bacteria found in the air 
of sewers. The information that has been derived from studies made 
upon such air relates to the number of bacteria present, and this, as 
already stated, is demonstrated to be usually lower than that for the CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
45 
open streets and inhabited houses, and we have no trustworthy evi- 
dence that pathogenic bacteria have ever been detected in sewer air. 
Notwithstanding this, no less of an authority on sewage than Mr. 
Corfield,22 as late as 1892, writes that, in addition to typhoid fever, 
proof has been adduced that diphtheria, erysipelas, pyaemia, septicae- 
mia, and puerperal fever occasionally result from exposure to such 
air. If this be true, an agent possessed of such manifold potencies 
for the production of disease should certainly call for much more 
careful study by exact methods than it has thus far received. The 
teaching that, because materials containing organisms capable of pro- 
ducing disease are constantly gaining access to sewers, that the air of 
these sewers of necessity contains such organisms also, is simply an 
opinion ; it is not supported by observations that lead us to accept it 
as fact. With the demonstration by Naegeli23 that bacteria do not 
arise from unmolested moist surfaces, or those in which agitation is 
moderate, though decomposition may be in active progress, and the 
confirmation of this observation by Pumpelly24 and many others, the 
opinion was suggested that perhaps bubbling and splashing that may 
occur in putrefying solutions might be instrumental in wafting bac- 
teria from such solutions into the superincumbent air. This opinion 
was very largely formulated from the results of experiments by Pro- 
fessor Frankland25 who caused active bubbling to occur in solutions of 
lithium chloride by generating carbonic acid in them and testing the 
superincumbent air spectroscopically for lithium. By this method he 
was able to detect the presence of lithium in the air at the mouth of 
a conducting tube 21 feet long. He found also that the lithium par- 
ticles passed through a layer of coarse charcoal five inches thick, 
placed in the tube. From these studies he concludes " that the break- 
ing of minute gas bubbles on the surface of a liquid .... is a potent 
cause of the suspension of transportable liquid particles in the sur- 
rounding air ; and, therefore, when through the stagnation of sewage 
or constructive defects which allow of the retention of excrementitious 
matters for several days in the sewer, putrefaction sets in and causes 
the generation of gas, suspension of zymotic matters in the air of the 
sewer is extremely likely to occur." 
The experiments of Pumpelly demonstrated that air aspirated 
through decomposing fluids at so low a velocity as % of a litre per 
day always carried bacteria with it into another flask with which it 
was in communication. 
Mr. A. E. Truby, student of bacteriology in our laboratory, has 
made a series of experiments under my direction on this point, the 
results of which, while in the main confirmatory of the experiments 46 
SEWER gas: 
of preceding investigators, afford some additional light upon the sub- 
ject. It was found, as others before had found, that with artificially- 
produced bubbling, resulting from the aspiration of air through solu- 
tions containing bacteria, that a certain number of these are thrown 
into the superincumbent atmosphere, and some of them are carried in 
the direction of the main air current. The number of bacteria car- 
ried is dependent upon several factors, the principle being the quan- 
tity of air aspirated and the velocity at which it is drawn through 
the decomposing matters. He found that when from 5 to 7 litres of 
air were aspirated through solutions containing known species of bac- 
teria, that none of these bacteria were transported into a flask of 
sterilized medium placed 1 meter distant at the end of a connecting 
tube of 15 millimeters inside diameter when the velocity of the air 
current in this tube did not exceed 8.6 cm. per minute, but that if 
the velocity was doubled, bacteria were carried over this distance in 
every case. On the other hand, in five experiments in which the 
length of the tube connecting the decomposiug infusion and the flask 
of sterilized medium was increased to 4.5 meters, 7 litres of air with 
the velocity increased 30 fold were aspirated through without bacte- 
ria being carried this distance in three out of five cases (for details 
see Table VI). Do experiments of this nature in any way simulate 
conditions naturally found in sewers, and do results obtained from 
such experiments afford a trustworthy basis for the assumption that 
because of splashing and bubbling in the sewer, bacteria are of a 
necessity thrown into overlying air and conveyed through the manifold 
ramifying arms of the sewer into houses in connection with it ? To 
me it seems most unlikely. In the experiments to which reference is 
made there is always a single direct current of air passing through 
the infected solutions, causing more rapid and vigorous bubbling than 
I have ever been able to detect, even in the most actively decompos- 
ing infusions, and passing thence for a longer or shorter distance 
through straight or slightly bent smooth tubes directly into a medium 
that is to demonstrate the presence of bacteria that the air may con- 
tain. There is nothing unfavorable to the success of such a demon- 
stration when conducted within reasonable limits. 
With the hope of approaching more closely to natural conditions I 
arranged a number of experiments in a different way, the object 
being to determine if in a closed chamber from which air currents 
were excluded bacteria did arise from substances undergoing bubbling 
as a result of decomposition. To this end a large bell jar of 40 centi- 
meters diameter, 28 centimeters high and of approximately 37 litres 
capacity was employed. Under it was placed an open glass dish of CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
47 
about 11 centimeters in diameter and 7 centimeters high containing a 
fluid culture of bacteria, and arranged around this were Petri plates 
containing sterilized gelatine, from 4 to 7 having been used so that 
there was an exposure of from 310 to 550 square centimeters of nutri- 
ent gelatin upon which bacteria could fall from the air. Bubbling 
was brought about in two ways. First, spontaneously, by the addi- 
tion to the culture of from 4 to 5 per cent, of glucose and then inocu- 
lating it with ordinary baker's yeast, and second, by the propulsion 
of air through it by means of a glass tube passed through an opening 
in the side of the bell jar. In both cases the bubbles arose to the 
surface of the fluid, burst and directed whatever was disengaged from 
the fluid straight up, the direction not being influenced by steady 
currents in one or another direction to any greater extent than could 
be avoided. Because of the manifold chances of contamination dur- 
ing the manipulation of so much bulky apparatus, this is hardly an 
ideal bacteriological experiment, but the results that have been ob- 
tained are, I think, of interest, for I believe they approach more 
nearly to what we may expect under conditions that are found in nature. 
As a result we found that when a pure culture of an easily recog- 
nizable organism-the bacillus prodigiosus-was placed in a wide 
open vessel in the chamber and bubbling or foaming was induced 
through the action of yeast upon glucose, not a single colony of this 
organism appeared upon either of seven plates exposed in the chamber 
for three days. There did, however, appear a few colonies of other 
organisms, principally moulds, but similar results to this were obtained 
in a control experiment made in the same chamber from which the 
culture was absent, so that probably these colonies represented acci- 
dental contaminations. 
When, on the other hand, bubbling was produced by propelling air 
through a similar fluid culture, at even so low a velocity as 3| litres 
in 6 hours, the bubbles rising to the surface at the rate of 22 per 
minute, colonies of the organism were seen on the plates in varying 
numbers, from 12 to 38 per plate, on the following morning. As with 
the preceding experiment, a few foreign colonies were also present. 
From these observations it is clear that the number of bacteria 
thrown into the air as a result of bubbling from spontaneous fermen- 
tation is probably very much smaller than results obtained from arti- 
ficially produced bubbling would lead us to infer. Let us now com- 
pare these observations made in closed chambers under conditions 
within our control with the phenomena that are presumably occurring 
in sewers. In the first place, as ordinarily constructed, sewers do not 
occasion very much splashing of their contents except, perhaps, at 48 
SEWER gas: 
isolated points. Secondly, there is no doubt that bubbling does 
occur, but it is probably feeble, at any rate I have never seen it to be 
active in sewage confined in vessels under temperature favorable to 
the process. Thirdly, the direction of air currents in sewers, as we 
know from the observations of Rosaheigyi and of Crimp (1. c.), is 
more often toward the mouth of the sewer than in a reverse direction, 
and almost as frequently there is no current at all. But let us sup- 
pose that a current did exist and always in the direction of the house, 
does it seem reasonable to suppose that this air, relatively poor in 
bacteria of all kinds, as experiment has shown it to be, would be 
likely to convey pathogenic bacteria, the presence of which is doubt- 
ful, from the sewer into the small branches, from these into still 
smaller branches, all moist, all presenting deviations from the straight 
line, and thence through soil pipes into houses, through water traps, 
or what is considered more dangerous, pin hole openings, the work of 
time or the careless plumber ? The conditions that it would meet on 
its passage from the sewer into houses are not such as to favor an 
increase in the number of bacteria in it, for all experiments agree in 
demonstrating that such particles are not dislodged from moist sur- 
faces even by winds of more than ordinary velocity. On the con- 
trary, everything is favorable to a diminution in the number of organ- 
isms ; with every change in the direction of the pipes conditions 
present that are favorable to the deposition of suspended bacteria 
upon the opposing moist surfaces, or their gravitation to the bottom, 
just as dust particles become deposited under similar circumstances. 
On this point the work of Carnelley and Haldane28 demonstrated the 
interesting fact that in passing through a tube along the bottom of 
which water was flowing the air gradually lost a large proportion of 
its bacteria. As a mean of 7 experiments they found that in flowing 
as short a distance as 5 feet under these circumstances the number of 
bacteria in the air was reduced a little more than 50 per cent., and 
they conclude "that it seems almost unreasonable to suppose that 
micro-organisms can enter a house with the air of sewers, unless the 
draught is very strong and rapid." Carmichael27 has also conclusively 
demonstrated that the intervention of water traps is a reliable safe- 
guard against an invasion of bacteria, and that under ordinary cir- 
cumstances they permit of less than 0'01 of a grain of gaseous matters 
to pass through them. 
But why consume youi' time with arguments against the possibility 
of this air conveying bacteria from sewers into houses when there is 
nothing definite to lead us to believe that the bacteria of such air are 
pathogenic ? CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
49 
If the limited studies that have been made upon the destiny of 
pathogenic bacteria, when in competition with the more vigorous 
saprophytic forms, are capable of general application, we may rea- 
sonably assume that the conditions existing in sewers are hardly 
favorable to the prolongation of the life of these organisms when 
deposited in sewers. Surely if this is not the case, or if they are not 
rapidly swept away with the current of sewage, which seems most 
improbable, we should observe the much more frequent occurrence of 
acute infections among the workers in sewers than is indicated by 
the results of statistical studies on the subject. To meet this latter 
point it has been suggested that the workers in sewers acquire a toler- 
ance to the conditions by which they are surrounded; this is plausible 
in so far as the gaseous matters in the air are concerned, but I fancy 
no one would as yet feel justified in advancing this hypothesis with 
regard to the organized, infectious agents, if they were as numerous, 
and constantly present in the air of sewers as some would have us 
believe. 
The more conservative of those who are inclined to the belief in a 
causal relation between the air of sewers and pathological conditions 
are gradually coming to the opinion that it is not directly concerned 
in the production of disease, but that its continuous respiration in 
some way interferes with the normal, vital resistance of the tissues and 
thus renders them more susceptible to infections to which they may be 
exposed. In very recent times observations have been made that may 
prove to have a most important bearing upon this opinion. The 
studies of Sanarelli28 upon experimental typhoid fever resulted in the 
interesting demonstration that animals not normally susceptible to 
this disease could be rendered susceptible to inoculation with the 
bacillus of typhoid fever by previous injection into their tissues of 
the products of growth of several organisms not normally concerned 
in disease production, among these being the products of ordinary 
putrefactive bacteria. While it is true that such procedures do not 
simulate very closely the respiration of sewer air or air from other 
decomposing matters, still, they demonstrate that in concentrated 
form the products of decomposition have a decidedly depressing action 
upon the resistance of the animal into which they are introduced. 
Of still more importance in their bearing upon the subject are the 
observations of Alessi29. He found that animals not normally open to 
infection by the bacillus typhi abdominalis became so after varying 
periods of exposure to the air of cesspools, and that the susceptibility 
was highest during the first two weeks of exposure, after which the 50 
SEWER GAS: 
animals seemed to gradually gain a tolerance and were then less suscep- 
tible to infection. 
I have also endeavored to get some light on the subject by experi- 
ments on animals but my work on this line was begun so late that I 
am hardly prepared to do more than mention it. 
The experiments were arranged with the idea of determining if pro- 
longed respiration of the gaseous products of decomposition would 
have any effect upon the health of animals, or their resistance to 
infection. With this in view 4 animals (3 guinea pigs and 1 rabbit), 
were confined in glass chambers through which was aspirated air that 
had bubbled through decomposing infusions of meat; as this was the 
only air obtainable by the animals for respiration, and as the infusions 
through which it bubbled were advanced in putrefaction, it is fair to 
assume that the volatile products of decomposition reached the ani- 
mals in a tolerably concentrated form. Of the three guinea pigs thus 
confined one (Pig A), was inoculated subcutaneously with fluid cul- 
tures of the bacillus typhi abdominalis twice, and subsequently into the 
peritoneal cavity with a similai' culture twice. The first inoculation 
was subcutaneous and was made after the animal had been respiring 
the gaseous products of decomposition for 14 days ; the last was into 
the peritoneal cavity and was made 6 days before death. The animal 
lived 30 days from the beginning of the experiment, having been con- 
fined in the chamber for 16 days of this period. It diminished in 
weight during this time from 460 to 300 grams, and had at one period 
a sudden fall in temperature to 35° C., immediately following inocu- 
lation. During the balance of the time the temperature fluctuations 
were unimportant. 
At autopsy on this animal there was no evidence of an acute or 
general infection, bacteriological examination revealing the presence 
of typhoid bacilli in only the sites of subcutaneous inoculation, the 
liver and bile. The most important tissue changes were those presented 
by the kidneys. They were strikingly altered in their naked eye 
appearance, the surface was roughened by elevations and depressions, 
but the capsules was not adherent and stripped off clean and easily. 
The color was not markedly altered. 
Sections of these organs, hardened in alcohol, present features cor- 
responding with the gross appearances. The areas of depression on 
the surface correspond to foci of proliferating fixed connective tissue 
elements. A considerable number of small granulation cells and larger 
fibroblasts occupy these areas, obscuring many of the tubules and 
compressing others. The glomeruli were likewise surrounded by such 
cells and in addition, the capsule of Bowman was at many points much CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
51 
thickened. These areas, which were usually triangular in shape, with 
the bases at the capsule, were much contracted, the partially atrophied 
glomeruli being, by this means brought closer together. The thick- 
ening seen in the capsule of Bowman is due to the presence of an excess 
of fibrous tissue, poor in cellular elements. Adjoining the atrophic areas 
the tubules were seen to be dilated and the epithelium lining then was 
swollen and degenerated. In a few tubules hyaline casts were ob- 
served. In short, the condition was identical to that seen in the 
kidneys of human beings dead from chronic Bright's disease. (See 
photograph No. 1 and No. 2.) 
The next animal, Rabbit V, was confined in the same chamber and 
kept under similar respiratory conditions. It lived 21 days, during 
which interval fluid cultures or suspensions of the bacillus typhi abdom- 
inalis were injected into its circulation by way of the ear vein on 
three occasions, 1.5 c.c. having been used for each injection. The ani- 
mal died on the day following the last injection as did a control ani- 
mal not exposed to the gases of decomposition. 
During the 21 days that it was under observation its weight fell 
from 1080 to 995 grams and its temperature fluctuations were insig- 
nificant. At autopsy typhoid bacilli were recovered from the liver, 
and spleen ; the blood, urine and kidneys being negative. As with 
pig A the conspicuous alterations in this animal were those found in 
the kidneys. They presented upon the cortical surface a remarkable 
mottled appearance, being thickly studded with round, pale, flat areas 
that looked to the naked eye like minute cysts, about the size of a pin 
point. A number of small red points were also seen. The capsule 
stripped off easily and clean. On section the cortical portion is red, 
the medullary pale. Sections of these organs, hardened in absolute 
alcohol, revealed conditions suggestive of the action of a very energetic 
poison. Unlike the chronic conditions found in pig A, there was here 
an acute process. Everywhere throughout the entire organ the inter- 
tubular capillaries were literally plugged with polynuclear leucocytes, 
causing them to stand out in bold contrast with the tubular elements, 
particularly the convoluted tubules. The epithelium of the convoluted 
tubules was granular and necrotic, the nuclei did not take up the stain- 
ing and in a few of these tubules fibrin was demonstrated by the method 
of Weigert. It is interesting to note that this necrosis was especially 
seen in the epithelium of the convoluted tubules and did not extend 
to that of the collecting tubes, the epithelium of which showed but 
little deviation from the normal. At many points of contact this 
difference was conspicuous. The nuclei of both parenchymatous and 
interstitial elements frequently gave evidence of the action upon it of 52 
SEWER gas: 
a violent poison, by destructive changes of varying degree that were 
present at different points. The glomerular nuclei were conspicuously 
affected, being fragmented, distorted, pinched off into buds or drawn 
out into long thread-like processes and otherwise altered in shape. 
Many of the glomerular capillaries were hyaline and at some points 
there was an accumulation of granular matter within the capsule of 
Bowman. The entire process was one that suggested the activity of 
a violent toxic agent. (See photograph No. 3.) 
Control animals inoculated in different ways with cultures of the 
typhoid bacillus showed at autopsy no appreciable alterations in the 
kidneys. 
Two guinea pigs were then placed in the chamber and subjected to 
the gaseous products of decomposition alone; not being inoculated 
at all. 
The experiment was begun April 11th, 1894 and continued with no 
interruption for 27 days, when, on the night of the last day the aspira- 
tor ceased to work and both animals were asphyxiated during the night. 
At autopsy there was no macroscopic evidence of decomposition, 
nor did sections of the organs show such change. There was nothing 
abnormal about the kidneys of either animal to the naked eye. 
They were placed in absolute alcohol and in Fleming's solution. In 
addition to the changes that will be described from the alcohol prep- 
arations those from Fleming's solutions showed the presence of a few 
minute black points, doubtless fat, in the epithelium of the convoluted 
tubules. 
Further examination of sections from the kidneys of both animals 
showed a striking increase in the number of nuclei apparently between 
the tubules and surrounding the glomeruli in the cortex. The increase 
of nuclei was not equally present in all parts, nor was it present to the 
same degree in both animals. 
In pig K, in which the change was less marked, the increase was 
apparently limited to the blood vessels. Large areas of the sections 
often showed no deviation from the normal, whereas along the distri- 
bution of one or two intertubular arteries the increase of nuclei was 
apparent. For the most part the cells presented the appearance of 
fibroblasts, were possessed of epitheloid nuclei and in no instance 
could fibres be made out between them. Sometimes the increase was 
especially about the glomeruli, resulting in an apparent thickening of 
the capsule, at other times it was more marked between the tubules. 
In the case of pig J (the companion of the preceding animal), the 
changes resulting from cellular proliferation in the cortex are still more 
marked. (See photograph No. 4.) In the first place the areas of cellu- CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
53 
lar proliferation are much more numerous and the distribution is some- 
what different. Such localizations as have just been described in the 
companion animal (pig K) namely : the increase of cells along the 
cortical vessels, is found here also, but in addition there are small 
groups of cells just beneath the capsule, extending a short distance 
downwards into the substance of the kidney, and associated at these 
points with a slight indentation of the surface of the organ. In this 
animal, as little as in the other, could fibres be made out distinctly be- 
tween the cells. The Fleming preparations showed, as did those of 
the other animal, a small amount of fat in the epithelium of the con- 
voluted tubules. Beyond this it cannot be said that marked changes 
in the epithelium were demonstrable. A control guinea pig of the 
saihe lot, killed for purposes of comparison, showed in the kidney 
lesions that are similar to those described in the last two animals, 
although they are less marked than in the latter (pig J) and more 
marked than in the former (pig K). In fact they correspond so 
closely in many respects with the lesions seen in pig K that we feel 
further investigations upon new lots of animals to be necessary before 
definite conclusions can be drawn. If, however, one is justified in 
formulating an opinion from the results that have been obtained in 
these experiments, we think it evident that they suggest the existence 
in the concentrated emanations from decomposing matters of sub- 
stances which when respired, have the power of producing alterations 
in the integral elements of the kidneys of animals, but we do not, as 
yet, consider this point demonstrated. 
It is equally interesting that animals can live for so long a time in 
such an atmosphere and experience no greater pathological lesions 
than those to which allusion was made in guinea pigs J and K. If 
the results of experiments made upon animals in this way are of any 
value in demonstrating the positive or negative effect of air saturated 
with gaseous products of decomposition, it does not seem reasonable 
from these studies to suppose that the air of a sewer or cesspool, in 
the enormous dilutions that it exists at the time it reaches an individ- 
ual in a house to which it has access, can be of much importance in 
either the direct production of diseased conditions or, in alone influ- 
encing the vital powers of resistance of the individual who inhales it. 
Note.-Before closing I wish to acknowledge my indebtedness to 
Dr. Simon Flexner, Associate in Pathology, Johns Hopkins Univer- 
sity, for his valuable assistance in the study of the pathological con- 
ditions found in the animals upon which the above experiments were 
performed. 54 
SEWER gas: 
BIBLIOGRAPHY. 
1 de Clambray-See Parent Duchatelet. Hygiene Publique, etc. 1836, Tome 
I, pp. 389 and 340. 
2 Lethby-Report on Sewage and Sewer Gases, and on the Ventilation of 
Sewers. London, 1858. 
3 Nichols-"Chemical Examination of Sewer Air." Report of Supt. of 
Sewers, Boston, 1879. 
4 Miller-" On the Ventilation of Sewers." Chemical News (American re- 
print) Vol. II, 1868, p. 214. English edition. Vol. XVII, 1868, 
Meh. 13th. 
f Russell-"Experiments on the Air in Sewers and Drains." Rep't British 
Association for Advancement of Science, Liverpool, 1870, Sept. 
5th, pp. 72-74. 
6 Beetz-Aerztliches Intelligenz-blatt. No. 20, 15 Mai, 1877; "Ueber die 
Luft in Kanalen." 
7 Miquel-Les organismes vivantes de l'atmosphere, 1883. 
8 Petri-Zeit. fur Hygiene, Bd. HI, 1888. 
9 Carnelley and Haldane-Proc. Royal Society. Vol. 42. 
10 Chandler-The Sanitary Engineer. Feb. 16, 1882, p. 240. 
11 Rosahegyi-Zeit. fur Biologie. Bd. XVII. 1881, p. 23. 
12 Crimp-Experiments on the Movement of Sewer Air at Wimbledon, Min- 
utes of Proceedings of the Institution of Civil Engineers, Vol. 
XCVII, Pt. Ill, 1888-89. 
13 Roscoe and Lunt-" Contributions to the Chemical bacteriology of Sewage." 
Philos. Trans, of the Royal Soc. of London; 1891, Vol. 182, 
p. 633. 
14 Hesse-Zeit. fiir Hygiene. Bd. XV, 1893. 
16 Erismann-Zeit. fiir Biologie. Bd. XI, p. 207. 
16 Petri and Maasen-Arbeiten aus dem Kaiserlichen Gesundheitsamte. Bd. 
VIII, Heft. HI. 
17 Stagnitta-Balisteri. Arch, fiir Hygiene, Bd. XVI, 1893. 
18 Remsen-National Board of Health Bulletin, Washington, D. C. Sept. 11th, 
1880. 
19 Drown-Reports of the Massachusetts State Board of Health, 1890, Part I, 
p. 524. 
20 Hazen and Clark-American Chemical Journal, Vol. 12, No. 6. 
21 Carnelley and Mackie-Proc. Royal Soc. London, 1886, No. 41, page 238. 
22 Corfield-Treatise on Hygiene, by Stevenson and Murphy, Vol. 1, p. 839-840. 
23 Naegeli-"Die niederen Pilze." pp. 109-111. 
24 Pumpelly-Report of the National Board of Health, Washington, D. C. 
1881. 
25 Frankland-Proc. Royal Soc. London, Vol. XXV, 1876-77. 
26 Carnelley and Haldane-See 9. 
27 Carmichael-The Sanitary Journal, Glasgow. 1880, March 1st. n. s. No. 49. 
28 Sanarelli-Annales de 1'Institut Pasteur, Tome VI. 1892. 
29 Alessi-Centralblatt fiir Bacteriologie u. Parasiten-Kunde, Bd. XV, 1894, p. 
228. 
26 Carnelley and Haldane-See 9. CHEMICAL, PHYSICAL AND BACTERIOLOGICAL STUDIES. 
55 
PHOTOGRAPHS. 
No. 1-Section through kidney of Guinea Pig A; showing condition of 
chronic interstitial nephritis, x 125. 
No. 2-Section through another portion of same kidney, x 60. 
No. 3-Section through kidney of Rabbit V ; showing accumulation of poly- 
nuclear leucocytes in the intertubular capillaries and necrotic 
condition of tubular epithelium, x 125. 
No. 4-Section through kidney of Guinea Pig J; showing accumulation of 
fibroblasts at points about the glomeruli; and in some places 
apparent thickening of the capsule of Bowman, x 125. 
Note.-I am under many obligations to Doctor Walter Reed, Surgeon U. S. 
Army, Curator of the Army Medical Museum, Washington, D. C., for his 
kindness in having had these photographs made for me. Showing condition of the air over decomposing substances confined in closed vessels, after varying intervals. All main- 
tained at the ordinary temperature of the laboratory. 
Date. 
Material. 
Period of confine- 
ment in flask. 
Oxygen, as vol. 
per cent. 
Total gases ab- 
sorbed by KOH. 
Vol. per. cent. 
Remarks. 
Oct. 5, 1893. 
Sewage. 
12 days. 
17.25 
1.42 
No distinct odor, neutral reaction. 
Nov. 8, 1893. 
u 
35 " 
13.39 
4.12 
cc cc cc a a 
Feb. 12, 1894. 
u 
35 " 
6.50 
14.4 
Consists of slime, etc., scraped from the bottom 
of sewers and mixed with fluid sewage. 
April 23, 1894. 
u 
7 " 
0.45 
23.63 
Consists of slime, etc., scraped from the bottom 
of sewers and mixed with fluid sewage. 
Oct. 1, 1893. 
Meat infusion. 
8 days. 
0.27 
34.8 
These infusions consisted of about 50 grams of 
finely chopped lean meat to a litre of water. 
Nov. 4, 1893. 
Ci 
15 " 
1.54 
19.28 
All samples, including the sewage samples, 
were kept in hermetically sealed flasks and 
April 23, 1894. 
a 
7 " 
0.22 
29.38 
were only opened for analysis. They were 
all maintained at the ordinary temperature 
of the laboratory. 
TABLE I. Analyses of air passing continuously over decomposing infusion of meat 
confined in flask. 
Table II 
Oct. '93 
23 
24 
25 
26 
27 
28 , 
29 
30 
31 
Nov. 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
Dec. 
2 
Date. 
58.3 
58.3 
69.6 
72.0 
90.0 
69.6 
90.5 
90.5 
186.5 
193.9 
110.6 
148.8 
148.9 
66.5 
74.1 
4.26 
5.16 
4.08 
6.10 
4.30 
5.54 
5.41 
58.0 
53.6 
101.5 
87. 
147. 
73.8 
55.6 
128.0 
129.4 
a=33 
^=9 
a 9 
a 9 
Litres of air 
drawn through 
during preced- 
ing day 
19.4 
19.4 
23.2 
24.0 
30.0 
23.2 
30.2 
30.2 
62.2 
64.6 
36.9 
49.6 
38.3 
22.2 
24.7 
1.42 
1.72 
1.36 
2.03 
1.43 
1.85 
1.80 
19.3 > 
17.9 
33.8 
29.0 
49.0 
24.6 
18.5 
43.0 
43.1 
11 
9 
3 
3 
Rate of renewal 
of air in flask 
for preceding 
day. 
82.2 
82.2 
156.0 
182.5 
376.2 
183.0 
185.3 
185.3 
326.8 
423.4 
602.8 
491.2 
465.6 
176.7 
526.2 
279.1 
398.5 
312.8 
374.2 
263.0 
244.1 
251.5 
449.7 
145.6 
270.5 
214.5 
246.5 
152.7 
237.3 
275.9 
236.2 
233.0 
113.0 
425. 
117. I 
Total CO2 ex- 
pressed as 
milligrams. 
J 35.8 
1 
- 
J 12.6 
16.2 
11.0 
6.1 
2.4 
69.0 
0.00 
1.1 
0.3 
0.00 
0.00 
8.7 
4.7 
0.4 
0.00 
0.1 
0.8 
3.60 
5.1 
0.00 
5.5 
7.1 
13.2 
1.4 
4.5 
1.8 
5.0 
Total H2S as 
milligrams. 
L4 
1.4 
2.24 
2.53 
4.18 
2.63 
2.05 
2.05 
1.75 
2.18 
5.45 
3.30 
3.13 
2.66 
7.11 
65.5 
77.2 
76.7 
61.4 
61.2 
44.1 
46.5 
7.8 
2.72 
2.67 
2.47 
1.68 
2.07 
4.27 
2.16 
1.83 
Relative CO2 as 
milligrams 
per litre. 
0.103 
0.029 
0.084 
0.099 
0.041 
0.016 
1.038 
0.00 
0.258 
0.058 
0.00 
0.00 
2.023 
0.848 
0.074 
0.00 
0.002 
0.008 
0.042 
0.035 
0.00 
0.099 
0.06 
0.102 
Relative H 8 S as 
milligrams 
per litre. 
23 
20 
20.5 
16 
' 21 
26 
21 
17.5 
13 
24 
23 
20 
22 
21.5 
23 
10 
15.5 
24 
16 
8 
25 
26 
22 
24 
21 
25 
26 
25 
25 
22 
15 i 
Temperature 
fluctuations in 
degrees C. 
§ • g g ® b ° 
® S ® S-' 2 B? B-' S- S $ 
B 92, ® omS obS ob|8 
gog 1 
• p 92? a 2. b - 
?« f ? sr °F ? £ g. Analyses of air passing continuously over sewage confined in flask 
Table III. 
Dec'93. 
11 
12 
13 
14 
15 
16 
17 
18 
19 
20 
21 
22 
23 
24 
25 
26 
27 
28 
29 
30 
Date. 
91.6 
84.8 
89.5 
82.0 
124.5 
110.5 
242.0 
167.9 
177.7 
163. 
80.5 
129.9 
73.9 
145.0 
123.0 
245.3 
37.5 
170.8 
Litres of air 
drawn through 
during preced- 
ing day. 
30.5 
28.3 
29.5 
27.3 
41.5 
36.8 
80.6 
55.9 
59.2 
54.3 
26.8 
43.3 
24.6 
48.3 
41.0 
81.8 
12.5 
56.9 
Rate of renewal 
of air in flask 
for preceding 
day. 
108. 
92.4 
123.2 
94.7 
114.0 
103.7 
106.2 
57.6 
99.3 
58.2 
105.5 
83.4 
51.4 
135.5 
44.3 
26.0 
53.0 
Total CO2 ex- 
pressed as 
milligrams. 
2.3 
2.2 
7.5 
0.00 
61.5 
12.9 
28.7 
12.9 
3.1 
0.00 
3.8 
16.3 
16.7 
72.5 
18.6 
21.2 
71.5 
Total H8S ex- 
pressed as 
milligrams. 
1.2 
1.1 
1.4 
1.2 
0.92 
0.94 
0.44 
0.34 
0.56 
0.36 
1.31 
0.642 
0.70 
0.94 
0.36 
0.61 
0.31 
Relative CO2 as 
milligrams 
per litre. 
0.03 
0.03 
0.084 
0.00 
0.494 
0.120 
0.12 
0.08 
0.017 
0.00 
0.48 
0.13 
0.21 
0.50 
0.15 
0.57 
0.42 
Relative H2S as 
milligrams 
per litre. 
22 
24 
21 
20 
24 
26 
19 
18 
21.5 
22 
21 
23 
20 
12 
20 
20 
23 
18 
Temperature 
fluctuations in 
degrees C. 
1 1 
5.® g-® p 
ws-a Mimi- 
« P GO • ® 5 ® g. 
GO ® on *00 p 
• E £ ® ® ® 
Sample 2, from another sewer. 
Jan. '94 
10 
86.5 
28.8 
354.9 
1.3 
4.10 
0.015 
21.5 
11 
98.5 
32.8 
245.4 
0.7 
2.50 
0.007 
19 
12 
107.3 
35.8 
209.7 
0.00 
2.00 
0.00 
15 
13 
185.2 
61.7 
242.8 
0.00 
1.31 
0.00 
18 
14 
120. 
40. 
84.4 
0.6 
0.70 
0.005 
19-14 
( Sudden fall in temp, 
f Note effect on HOS. 
15 
58.2 
19.4 
35.4 
1.9 
0.61 
0.033 
22 
f After weighing on 15th, 
16 
65. 
22. 
77.5 
43.3 
1.21 
0.67 
24 
| a dense pellicle that had 
17 
181.6 
60.5 
136.4 
38.8 
0.751 
0.22 
23 
■{ formed on the surface 
of the sewage was 
18 
181.2 
60.4 
81.1 
59.0 
0.48 
0.33 
22.5 
| broken. Note effect on 
19 
156.3 
52.1 
116.4 
40.6 
0.75 
0.26 
24 
[HgS. See chart.  
- 
- 
Material tested. 
Amount. 
Amount of 
water 
Distillation 
with or 
Contents of each fraction of 50 c.c. of distillate. 
Total 
nh3 
Correction 
for NH3 in 
water 
employed. 
Correct'd 
total. 
employed. 
pumice 
stone. 
1st 
2d 
3d 
4th 
5th 
Jith 
7th 
8th 
found. 
A 
B 
0 
Ammonium 
chloride 
(NH4C1) 
66 
66 
I 
II 
I 
II 
I 
II 
J- =1 mg. NH3 
=5 mgs. NH3 
=1 mg. NH3 
• 500 c.c. 
750 c.c. 
■750 c.c. 
With 
With 
With 
0.900 
0.900 
3.250 
3.250 
0.750 
0.800 
0.175 
0.175 
1.300 
1.200 
0.250 
0.1900 
0.033 
0.040 
0.475 
0.5250 
0.090 
0.090 
0.010 
0.007 
0.170 
0.170 
0.038 
0.038 
0.002 
0.0025 
0.055 
0.055 
0.0105 
0.0105 
0.000 
0.000 
0.018 
0.014 
0.0025 
0.0025 
0.006 
0.0055 
0.000 
0.000 
0.002 
0.000 
0.000 
0.000 
1.120 
1.1245 
5.2810 
5.2195 
1.141 
1.131 
• 0.0185 
• 0.027 
■ 0.0425 
1.1015 
1.106 
5.254 
5.1925 
1.0985 
1.0885 
D 
E 
F 
G 
Ammonium 
chloride 
(NH4CI) 
66 
66 
6,6 
I 
II 
I 
II 
I 
II 
I 
II 
j- =5 mgs. NH3 
=1 mg. NH3 
=5 mgs. NH3 
J- =1 mg. NH3 
■ 500 c.c. 
• 500 c.c. 
• 750 c.c. 
750 c.c. 
Without 
Without 
Without 
Without 
3.250 
3.500 
0.830 
0.8300 
2.750 
2.900 
0.600 
0.650 
1.200 
1.150 
0.220 
0.1900 
1.200 
1.100 
0.325 
0.275 
0.3125 
0.300 
0.065 
0.045 
0.550 
0.575 
0.125 
0.125 
0.075 
0.060 
0.019 
0.013 
0.2438 
0.2375 
0.045 
0.055 
0.024" 
0.012 
0.0035 
0.0025 
0.0875 
0.0563 
0.014 
0.0225 
0.006 
0.006 
0.003 
0.000 
0.045 
0.0365 
0.005 
0.012 
0.0015 
0.000 
0.016 
0.011 
0.0025 
0.0025 
0.000 
0.000 
0.0045 
0.0035 
0.002 
0.0015 
5.069 
5.028 
1.1405 
1.0805 
4.8968 
4.9198 
1.1185 
1.1435 
■ 0.0295 
■ 0.0095 
0.0145 
• 0.0155 
5.0395 
4.9985 
1.131 
1.071 
4.8823 
4.9053 
1.103 
1.128 
H. 
I 
Trimethyl- 
amine. 
Methylamine. 
Methylamine. 
Trimethyl- 
amine. 
Mixture of 
Methylamine. 
Trimethyl- 
amine and 
Ammonium 
chloride 
5 c.c. of about 
1 per cent. sol. 
5 c.c. of about 
1 per cent. sol. 
20 c.c. of about 
1 per cent. sol. 
20 c.c. of about 
1 per cent. sol. 
5 c.c. of 1 per 
cent. sol. 
5 c.c. of 1 per 
cent. sol. 
2} mgs. NH3 
500 c.c. 
500 c.c. 
500 c.c. 
■ 500 c.c. 
>500 c.c. 
With 
With 
With 
With 
With 
1.150 
1.050 
3.500 
4.8260 
4.400 
0.2875 
0.240 
1.050 
1.300 
1.050 
0.085 
0.070 
0.2375 
0.2933 
0.2850 
0.017 
0.014 
0.0650 
0.055 
0.070 
0.003 
0.000 
0.0140 
0.0095 
0.013 
0.000 
0.000 
0.0035 
0.000 
0.0045 
1.5425 
1.3740 
4.870 
6.5038 
5.8225 
TABLE IV. 
Fractional distillations at 100° C., to determine the rate at which free ammonia and substitution products of ammonia pass off from solutions containing them. (1) 
Date. 
(2) 
Source of air 
analyzed. 
(3) 
Amount of 
air taken, 
in litres. 
(4) 
Time consumed 
(in hours) In 
aspirating. 
(5) 
Total ammonia 
found. 
(6) 
Ammonia in 
water used. 
To be deducted. 
Amount of 
ammonia in 
ail- analyzed. 
Remarks. 
Free. 
Alb. 
Free. 
Alb. 
Free. 
Alb. 
Dec. 23,1893 
Outside air. 
129.9 
24 
0.018 
0.0025 
0.0145 
0.0025 
0.0035 
0.000 
1 
Dec. 23,1893 
U CC 
35.2 
25.5 
0.0155 
0.0025 
0.0145 
0.0025 
0.001 
0.000 
Dec. 24,1893 
CC CC 
230.8 
49 
0.0145 
0.0025 
0.0145 
0.0025 
0.000 
0.000 
The air obtained for these analyses came 
Dec. 24,1893 
a cc 
47.6 
50.5 
0.0145 
0.0025 
0.0145 
0.0025 
0.000 
0.000 
1 from just outside a window of the laboratory, 
Dec. 26,1893 
cc cc 
355.4 
48 
0.0095 
0.0018 
0.006 
0.0018 
0.0035 
0.000 
on the second floor, about 35 or 40 feet from 
Dec. 26,1893 
CC U 
108 
46.5 
0.0075 
0.0025 
0.005 
0.0025 
0.0025 
0.000 
the ground. 
Dec. 28,1893 
u cc 
56.1 
20.5 
0.0075 
0.0025 
0.005 
0.0025 
0.0025 
0.000 
Dec. 28,1893 
CC cc 
182.2 
25.5 
0.0073 
0.0025 
0.005 
0.0025 
0.0023 
0.000 
J 
Dec. 13,1893 
Air of laboratory 
31.4 
2.5 
0.009 
0.009 
0.009 
0.009 
0.000 
0.000 
- 
Dec. 21,1893 
u cc 
105.3 
22 
0.0155 
0.003 
0.0145 
0.003 
0.001 
0.000 
Dec. 22,1893 
cc cc 
76.3 
4.5 
0.0175 
0.003 
0.0145 
0.003 
0.003 
0.000 
. The air used for these analyses was from 
Dec. 29,1893 
cc cc 
169.5 
22.5 
0.008 
0.0025 
0.008 
0.0025 
0.000 
0.000 
। my own room in the laboratory. 
Dec. 29,1893 
cc cc 
207.1 
21.5 
0.009 
0.002 
0.0085 
0.002 
0.0005 
0.000 
J 
Dec. 30,1893 
Air from sewage. 
36.9 
5 
0.036 
0.0025 
0.008 
0.0025 
0.028 
0.000 
In these experiments the air was aspirated 
Dec. 30,1893 
CC cc 
141.2 
5 
0.008 
0.004 
0.0085 
0.002 
0.000 
0.002 
over sewage contained in a flask. About four 
Jan. 3,1894 
cc cc 
36.1 
4 
0.072 
0.0065 
0.008 
0.0025 
0.064 
0.004 
'litres of sewage was placed in an eight litre 
Jan. 3,1894 
cc cc 
108 
4.5 
0.010 
0.004 
0.008 
0.0025 
0.002 
0.0015 
flask for the experiments. 
Jan.13,1894 
cc cc 
254.7 
42.5 
0.042 
0.0023 
0.0165 
0.0023 
0.0255 
0.000 
TABLE V. 
Estimation of Free and Albuminoid Ammonia in Air from different sources. Results expressed as fractions of a milligram. (1) 
Date. 
(2) 
Source of air 
analyzed. 
(8) 
Amount of 
air taken, 
in litres. 
(4) 
Time consumed 
(in hours) in 
aspirating. 
(5) 
Total ammonia 
found. 
(6) 
Ammonia in 
water used. 
To be deducted, 
(7) 
Amount of 
ammonia in 
air analyzed, 
Remarks. 
Free. 
Alb. 
Free. 
Alb. 
Free. 
Alb. 
Jan. 3,1894 
Jan. 4,1894 
Jan. 4,1894 
Jan. 7,1894 
Jan. 7,1894 
Air from sewage. 
CC cc 
cc cc 
cc cc 
CC CC 
61.9 
320 
866 
284.5 
387.7 
19 
51 
48 
22 
23 
0.0525 
0-4265 
0.590 
0.1755 
0.2740 
0.0025 
0.0045 
0.0075 
0.0039 
0.0038 
0.008 
0.0095 
0.0095 
0.0143 
0.0143 
0.0025 
0.0025 
0.0025 
0.0038 
0.0038 
0.0445 
0.4170 
0.5805 
0.1612 
0.2597 
0.000 
0.002 
0.005 
0.0001 
0.000 
1 
. In these experiments the air bubbled through 
1 the sewage before passing to the absorber. 
J 
Jan. 9,1894 
Jan.11,1894 
Jan. 12,1894 
Jan. 15,1894 
Air from decom- 
pos. meat infusi'n 
Air from decom- 
pos. meat infusi'n 
Air from decom- 
pos. meat infusi'n 
Air from decom- 
pos. meat infusi'n 
337 
192.4 
203.8 
178.3 
84 Hi 
80 in 
24 n 
I 
24 II 
III 
4.9815 
2.179 
1.205 
1.150 
2.521 
1.8815 
4.821 
1.806 
0.030 
0.0115 
0.014 
0.0025 
0.0025 
0.003 
0.0045 
0.0115 
0.004 
0.005 
0.009 
0.009 
0.012 
0.012 
0.0165 
0.0165 
0.0155 
0.0155 
0.0155 
0.0038 
0.0038 
0.0012 
0.0012 
0.0023 
0.0023 
0.0023 
0.0023 
0.0023 
4.9725 
2.170 
1.193 
1.138 
2.5045 
1.865 
4.8055 
1.7905 
0.0145 
0.007 
0.0102 
0.0013 
0.0013 
0.0007 
0.0022 
0.0092 
0.0017 
0.0027 
=absorber I 
" II 
=absorber I 
" II 
=absorber I 
" II 
=absorber I 
" II 
" III 
In these experiments the air was 
। drawn through vessels contain- 
ing decomposing infusions of 
meat. It did not bubble through 
the fluid but passed over the 
surface. For, each determina- 
tion a group of more than one 
absorber was employed. 
Jan. 8,1894 
Jan. 16,1894 
Feb.17,1894 
Feb. 18,1894 
Air bubbli'g thro' 
decomposi'g meat 
infusion, 12 ds. old 
Air drawn over 
sewage. 
Air bubbling 
through sewage. 
Air bubbling 
through sewage. 
143,2 
124.5 
200 
259 
I 
21 
II 
I 
20 II 
III 
21 n 
T 
25.5 TT 
1.549 
0.1715 
0.040 
0.0175 
0.0145 
0.0620 
0.0155 
0.323 
0.012 
0.0055 
0.004 
0.0015 
0.0013 
0.001 
0.000 
0.000 
0.002 
0.000 
0.009 
0.009 
| 0.0165 
[ 0.0095 
J 0.0095 
0.0038 
0.0038 
0.0025 
0.000 
0.000 
1.540 
0.1625 
0.0235 
0.001 
0.000 
0.0525 
0.006 
0.3135 
0.0025 
0.0017 
0.0002 
0.000 
0.000 
0.000 
0.000 
0.000 
0.002 
0.000 
=absorber j 
" II 
=absorber I 
" II 
" III 
absorber I 
" II 
absorber I 
" II 
TABLE V-Continued. 
Estimation of Free and Albuminoid Ammonia in Air from different sources. Results expressed as fractions of a milligram.  
Length of tubes in 
centimeters. 
Distance 
from 
culture to 
sterilized 
medium 
in cm. 
Calibre of tubes in 
millimeters. 
Litres 
of air 
aspira- 
ted. 
Time con- 
sumed in 
Velocities of air current 
in tubes, as cm.permin. 
Remarks. Results. 
Main. 
Connections. 
Main. 
Connections. 
aspirating. 
Main. 
Connections. 
1 
90 
40 
130 
15 
5 
7 
45 minutes 
88 
793 
Bacteria carried over, 6 colonies. 
2 
60 
40 
100 
15 
5 
7 
45 minutes 
88 
793 
" " " 5 
3 
140 
90 
230 
15 
5 
7 
90 minutes 
44 
397 
66 66 66 
4 
200 
40 
240 
15 
5 
7 
90 minutes 
44 
397 
66 66 66 
5 
90 
40 
130 
15 
5 
7 
120 minutes 
33 
297 
66 6 6 6 6 
6 
60 
40 
100 
15 
5 
7 
240 minutes 
16.5 
149 
6 6 6 6 6 
7 
140 
90 
230 
15 
5 
st 
360 minutes 
8.6 
78 
No bacteria carried over. 
8 
90 
40 
130 
15 
5 
7 
480 minutes 
8.3 
74 
66 66 66 66 
9 
90 
40 
130 
15 
5 
4 
360 minutes 
8.0 
57 
66 66 66 66 
10 
60 
40 
100 
15 
5 
5 
420 minutes 
4.8 
42.5 
66 66 66 66 
11 
- - - - 
448 
15 
-- 
7 
30 minutes 
134 
• - - - 
66 66 66 66 
12 
448 
15 
-- 
7 
15 minutes 
268 
66 66 66 66 
13 
- - - - 
448 
15 
-- 
7 
15 minutes 
268 
66 66 6 6 66 
14 
• - " - 
448 
15 
-- 
7 
15 minutes 
268 
Bacteria carried over. 
15 
---- 
.... 
448 
15 
-- 
7 
15 minutes 
268 
.... 
66 66 66 
TABLE VI. 
Transportation of bacteria by air bubbling through decomposing solutions. CHART I 
Showing tension of air over sewage as compared with that of air over decomposing infusion of meat. All kept under same conditions. 
' Tension of air over decomposing meat infusion, expressed in millimeters of mercury. Infusion confined in hermetically sealed flask. Experiment of Sept. 22nd, 1893. 
- " " " " Sewage from South St. sewer " " " " " Sewage " " " " " Experiment of Sept. 26th, 1893. 
Fluctuations of temperature in degrees Centigrade. 
-o-o- Tension of air over mixture of fluid sewage and slime scraped from sides and bottom of sewer. Mixture confined in hermetically sealed flask. 
-o-o-o- Temperature curve for preceding experiment. 
September, 1893. October. 
Fluctuations in pressure in millimeters of mercury. 
Fluctuations in tempera- 
ture in degrees centigrade. Showing absolute and relative amounts of Carbonic Acid and Sulphuretted Hydrogen taken up by air in passing over an actively decomposing infusion of meat. 
CHART II 
KEY : 
Rate of ventilation expressed as the number of times the entire volume of air in the flask (3 liters) was renewed during the preceding 24 hours. 
Total amount of CO2 given off during preceding 24 hours ; expressed in milligrams. 
Relative amount of CO2 expressed as milligrams per liter ; one milligram per liter at 20° C. being equivalent to about 10.7 parts per 10,000 of air (per volume). 
Total amount of H2S given off during preceding 24 hours ; expressed in milligrams. 
Relative amount of H2S expressed as milligrams per liter ; one milligram per liter at 20° C. being equivalent to about 7.34 parts per 10,000 of air (per volume) 
Fluctuations of temperature, expressed in degrees Centigrade. 
October, 1893. 
November. 
December. CHART III 
Showing absolute and relative amounts of carbonic acid and sulphuretted hydrogen taken up by air in passing over sewage confined in flasks at the ordinary temperature of the room. 
KEY : 
Rate of ventilation expressed as the number of times the entire volume of air in the flask (3 litres) was renewed during the preceding 24 hours. 
Total amount of CO2 given off during preceding 24 hours ; expressed in milligrams. 
Relative amount of CO2 expressed as milligrams per liter ; one milligram per liter at 20° C. being equivalent to about 10.7 parts per 10,000 of air (per volume). 
Total amount of H2S given off during preceding 24 hours, expressed in milligrams. 
Relative amount of H2S expressed as milligrams per litre ; one milligram per litre at 20° C. being equivalent to about 7.34 parts per 10,000 of air (per volume). 
Fluctuations in temperature, expressed in degrees Centigrade. No. 1. x 125, No. 2. x 60 No. 3. x 125, No. 4. x 125.