ON SOME POINTS OF INTEREST 
CONNECTED WITH THE WANKLYN METHOD OF 
SANITARY WATER ANALYSIS 
PARTICULARLY OK THE DETECTION OF 
RECENT SEWAGE 
AND THE DETERMINATION OF THE 
MATURE OP THE ORGANIC MATTER. 
BY CHARLES SMART, 
Major a,nd Surgeon U. S. Army. 
Reprinted from public health in Minnesota. 
Vol. 11., No. 3, May, 1886. NOT a great many years ago sanitary chemists affected a more precise 
knowledge of the organic matter in a water than those of the present day 
attempt to claim. Results were announced in grains per gallon; and the outside 
world conceived that the whole matter was understood. More recently the estab- 
lishment of arbitrary scales of purity or impurity, based on the results of special 
processes, has taken the place of the former method of giving a verdict, and the 
outside world, when not trammelled by personal considerations, has acquiesced 
in this assumption of special knowledge. Occasionally, as a consequence of 
this, an analyst at the present time falls in the esteem of certain individuals 
when he acknowledges his inability to give full information concerning the or- 
ganic matter in a water, to state positively whether a water is wholesome or un- 
wholesome, or to detect a trace of sewage which is known, or strongly suspected, 
to be present. Even men well versed in the general principles of sanitary 
science, have, by these acknowledgments, been led to regard the laboratory 
work of the analyst as a kind of quasi-scientific hocus-pocus, which tells as 
little as the ancient oracle. The fact is that although the principles involved 
in these processes are incapable of indicating whether a specific morbific 
agency is present, they very frequently throw much light on the probable char- 
acter of a water. Microscopical and biological inquiry will perhaps ultimately 
supercede the purely chemical methods; but the former are in their infancy as 
yet, and although very promising as children, it remains to be seen whether 
they will fulfill their promise. Meanwhile the work of the world in this line 
has to be performed by the chemical processes and it is encouraging to know 
that the more these are studied the more definite becomes the testimony they 
are capable of yielding. 
I propose here to refer in particular to certain results of the Wanklyn pro- 
cess that are not generally appreciated. This process measures the free 
ammonia in a water and afterwards the albuminoid ammonia, or that formed 
by the breaking up of the nitrogenous organic matters. The former, when in 
relatively large quantity, and particularly in well waters, is usually due to cess- 
pool, sewer, or privy connections. The latter is a measure of the existing im- 
purity, and, if relatively large, is naturally suggestive of unwholesomeness. 
Wanklyn’s arbitrary limits of wholesomeness are well known and need not be 
repeated. 
It is generally understood or assumed that the results given by the same 
water are always identical. On the contrary they may be modified by the man- 
ner in which the process is conducted. 
IMPERFECT CONDENSATION 
of the evolved ammonia causes a loss which may vary from 1 to 12 
per cent, of the total, according to the slowness or rapidity of the distillation and the temperature of the current in the condenser. Slow distillation and a 
condensing current at winter temperatures give the minimum of loss; rapid 
distillation and a condensing current at summer temperatures, the max- 
imum. This affects the total of both the free and the albuminoid ammonias; 
but the latter is affected even to a greater extent by the time occupied in the 
distillation, or by the 
IMPERFECT CONCLUSION OP THE EXPERIMENT. 
Wanklyn has acknowledged that his process is imperfect and that the 
whole of the nitrogen of the organic matter is not converted into ammonia; but 
he claims that, as the albuminoids in a water are of similar constitution, and 
yield up a definite percentage of their nitrogen, the results of the process in 
different instances are susceptible of comparison, and enable the operator to 
rate the water on an arbitrary scale. This claim is not wholly sustained by the 
results of recent investigations. The albuminoid ammonia does not pre-exist 
in the water; it is formed during the analytical operation, and requires time for 
its formation. The element of time is an important one. The nitrogen of 
some organic matters is slowly converted by boiling with the alkaline permanga- 
nate—so slowly that the experiment may reach its conclusion, by the disap- 
pearance of the water from the retort, before the whole of the nitrogen has been 
liberated as ammonia. This occurs with most of the alkaloidal substances. If 
an organic matter require boiling for two hours with alkaline permanganate to 
reduce and liberate its nitrogen a rapid progress of the experiment concluding 
the distillation in one hour will yield only a percentage of the nitrogen. A 
weak flame will give higher issues of albuminoid ammonia from the same or- 
ganic matter than a strong flame, because it prolongs the action on the one 
hand, and lessens the loss from imperfect condensation on the other. In all 
instances in which the organic matter is slowly decomposed, the yield of albu- 
minoid ammonia will vary with the time occupied by the distillation. Some 
waters will give very different results, according as the process is conducted at 
a rate of ten or of twenty minutes for each measure of 50 cubic centimeters dis- 
tilled. It is needful to bear these irregularities and imperfections in mind, for 
some of them give important indications as to the source and quality of the 
organic matter. 
DETECTION OE RECENT SEWAGE. 
This may be effected by noting the peculiar behavior of urea when 
submitted to the Wanklyn process. A few nitrogenous substances 
are not decomposed by the action of the alkaline permanganate solu- 
tion; among these Prof. Wanklyn includes urea. He is very positive 
on this point. He says: “Except in the instance of nitro-compounds, 
urea and ferrocyanide of potassium, we have not met with any unequivocal in- 
stance of failure of an organic nitrogenous compound to evolve ammonia in 
being heated to 100° 0, with a strongly alkaline solution of permanganate.” 
And again: “In presence of permanganate and excess of potash, urea is 
doubtless decomposed; but it yields no ammonia, which is a very extraordinary 
and noteworthy fact.” And further: “On inquiring into the other peculiarities 
of structure which prevent alkaline permanganate evolving nitrogen of a given 
organic compound in the form of ammonia our attention is arrested by the 
example of urea which evolves none of its nitrogen as ammonia when so 
treated.” Nevertheless, urea yields about 22 per cent, of its nitrogen as ammonia 
when treated as Wanklyn treated the alkaloids and other organic substances 
in determining the amount of their nitrogen evolved as ammonia by permanga- 
nate; and if the conditions of the experiment be arranged to permit of a longer 
continuance of the action of the permanganate on the urea, the whole of its 
nitrogen will be accounted for, provided that due allowance be made for loss 
from imperfect condensation. 
Urea in solution is resolved gradually into ammonia. This is hastened by 
boiling, and hastened yet more by boiling with alkaline permanganate. If a 
solution containing 1 mgrm. of urea in 500 c. e. of ammonia-free water be dis- 
tilled, the first measure of 50 c. c. which comes over will contain a compara- 
tively large quantity of ammonia, because during the time occupied by the 
liquid in reaching the point of ebulition a certain portion of the urea has been 
decomposed. The second measure will contain less, and the third, fourth and 
subsequent measures will each contain .01 mgrm. of ammonia, showing that 
boiling for the time ordinarily occupied in the distillation of 50 c. c. liberates 
the nitrogen necessary to form that amount of ammonia. Again, if a solution 
containing 1 mgrm. of urea and a charge of alkaline permanganate in 500 c. c. 
of ammonia-free water be distilled, the first measure of 50 c. c. which comes 
over will contain a relatively large quantity of ammonia. The second will con- 
tain less, and the third, fourth and succeeding measures will each contain .02 
mgrm. of ammonia, showing that the presence of the alkaline permanganate has 
as strong an influence in effecting the decomposition of urea as is exercised by 
the boiling temperature. After all pre-formed ammonia has been liberated from 
an ureal solution an equable and persisting evolution takes place which may be 
represented by N in the absence of caustic alkali and permanganate, and by 
twice N in the presence of these reagents. 
Dilutions of fresh or decomposing urine in tap water give similar results. 
When these are treated as water samples by the Wanklyn process there is first 
a liberation of the pre-formed ammonia, and afterwards an equable decompo- 
sition of the urea by the continued boiling. Then on the addition of the alka- 
line permanganate there is a comparatively free liberation of ammonia from the 
urea and other nitrogenous substances that may be present and afterward a 
persisting and equable evolution, each measure of 50 c. c. containing as much 
again as the corresponding measure distilled in the absence of the permanga- 
nate. This peculiarity in the behavior of urea is of importance, as by it the 
presence of this substance may be recognized in a water. It does not appear that 
any other substance gives issue to the free and the albuminoid ammonias in the 
manner stated. Many organic chemicals have been examined and also many 
organic substances, including the waste products of manufactures, etc., but not 
one was found to present reactions by which it could be comfounded with urea. 
Many gave a persisting and equable evolution of albuminoid ammonia, but few 
of these gave at the same time a persisting evolution of free ammonia; and 
where in some exceptional instances of factory drainings there was a persisting 
evolution of both free and albuminoid ammonia, these ammonias were not 
yielded in the proportion of 1:2. 
Moreover the observations on the decomposition of urea have an approxi- 
matively quantitative value;—for since 1 mgrm. of urea in 500 c, c. of water 
gives a persisting and equable evolution of .01 mgrm. of ammonia when distilled 
alone or with sodium carbonate, as in the first half of the Wanklyn process, and an evolution of .02 in each measure when subsequently treated with the alkaline 
permanganate, a water sample which yields such results must have contained 
urine equivalent to, at least, 1 mgrm. of urea. Thus, a water which yielded in 
successive distillates of 50 c. c., each .47, .25, .15 and .15 mgrm. of free ammonia, 
and afterward .54, .34, .32 and .32 mgrm. of albuminoid ammonia, respectively in 
each of four measures distilled from alkaline permanganate, might be set down 
as having contained urine equivalent to, at least, 15 mgrms. of urea in the 500 
c. c. of water used in the experiment. The urea in urine is, of course, a varia- 
ble quantity, but as the mean of a number of experiments on fresh urine, it was 
found that one half of a cubic centimeter in 500 c. c. of distilled water, 
or 1 part in 1000, yielded the returns given above as indicating the presence of 
15 mgrm. of urea in the 500 c. c. treated by the Wanklyn operation. From 
these data the quantity of urine present in a given sample of polluted water 
may be approximately estimated. 
The writer was so impressed with the importance and reliability of these 
observations on the persisting evolution of free and albuminoid ammonia from 
urea that on one occasion when an unknown water, sent to him for examina- 
tion, yielded such results, he did not hesitate to report the water as polluted by 
so much urine per gallon. His purpose was to make this a test case; and he 
succeeded, although he was not informed of the fact until nearly two months 
afterward, when, having reported another unknown water as a rain water, shed 
from a clean roof and stored in a sound and clean cistern, he was made ac- 
quainted with the following particulars:— 
Typhoid fever appeared in a cottage built by a gentleman as a summer 
residence. The quality of the water was suspected, and a sample analyzed 
by Prof. E. 8. Wood was pronounced unfit for use. A second chemist hav- 
ing confirmed this verdict, a physician inspected the premises and suggested 
that sewer gases from the ventilating pipe of the water closets might 
have been condensed on the roof and washed thence into the cistern. Meas- 
ures were taken to remedy this assumed evil, and the cistern was pumped out, 
cleaned and relined with cement. The water sample which was condemned 
as containing so much urine per gallon, was taken from this cistern after it had 
thus been subjected to careful and scientific treatment. Naturally, the proprie- 
tor was shocked; and, in view of his efforts to procure a fair rain water, felt 
disposed to question the analytical results rather than the quality of the water. 
Fortunately, however, Mr. E. W. Bowditch, of Boston, who was conducting 
the sanitary survey of the premises, conceived that further investigation was 
imperatively required. Ultimately in his examination he discovered that there 
were three apertures in the cistern, although only two conductors from the roof 
entered it. On inquiry it was remembered that a few years before, in making some 
additions to the cottage, a conductor had been disused, but what had been done 
with it was not known. The track of this old conductor was then uncovered at 
its cistern end, and followed to its termination in an open mouth a foot below 
the surface near the porch of the building. A luxuriant growth of vines rose 
from the surface at this point, and these vines were from time to time nourished 
with chamber slops. It was moreover found to be the custom of the house to 
collect all such slops in pails which were emptied through a water closet on the 
first floor and then set out on the roof of the piazza to air. The old conductor 
was removed and its cistern aperture sealed, and the connection was cut be- 
tween the cistern and the roof of the piazza. When the water again accumu- lated after these changes had been effected, the sample submitted for analysis 
was reported as a satisfactory rain water. This appears to illustrate the value 
attaching to the manner in which the ammonias come over during the distilla- 
tions of the Wanklyn process, and it is specially mentioned here on account of 
the difficulty encountered in verifying the accuracy of the chemical results. 
Many other illustrations might be given, but they are unnecessary. 
Wanklyn has observed that when a water containing vegetable mat- 
ter is distilled with the alkaline permanganate solution the albuminoid 
ammonia comes over very slowly. In verifying the accuracy of this 
statement, it was found that the rapidity of the evolution was determined 
by the instability of the matter, or its tendency to a state of putrescence 
rather than by its derivation. Pure animal and pure vegetable albumi- 
noids gave up their nitrogen as ammonia at an equal pace when equal times 
were occupied in the distillation of the various measures. Thus, if the first 
measure of the distillate contained .24 mgrm., the second would contain .12, the 
third .06, the fourth .03, etc. But if these matters were in a decomposing con- 
dition, the second, third, and subsequent measures would each contain only 
one-third of the ammonia found in the measure that immediately preceded it. 
Animal matters, however, more readily pass into the putrefactive condition, so 
that when the organic matter of a water gives up its nitrogen after the manner 
first stated, there is a strong probability of its derivation from the vegetable 
kingdom; and this probability is rendered almost a certainty if a comparatively 
large quantity of oxygen is required for its oxidation when treated by the Kubel 
or Tidy method. 
ANIMAL OB VEGETABLE MATTEE. 
But vegetable matter in a state of fermentative change reveals itself during 
the Wanklyn process in two ways, neither of which appears to have been ap- 
preciated by any of the many analysts who make use of this method of sanitary 
analysis. The color produced by the Nessler reagent in a weak ammoniacal 
solution varies from a pale straw color to a dark sherry brown; but occasionally 
in testing the distillates of the Wanklyn process, particularly those containing 
the free ammonia, a green coloration masks the brown of the ammonia and pre- 
vents the accurate estimation of the volatile alkali. If there be but a trace of 
ammonia present, the color may be an olive-green, or even a citron-green, with 
generally a tendency to the development of a haziness in the liquid. No ex- 
planation of this color-interference has been furnished, although some years 
ago a reference was made to it at a meeting of the English Society of Public 
Analysts. Wanklyn does not even mention its occurrence. 
The frequency with which this color-interference was encountered in cer- 
tain experiments on water washings of air drawn over fermenting vegetable re- 
fuse, led me to conceive that it was due to an ethylic compound formed dur- 
ing the fermentation. A reference to my laboratory note books speedily showed 
that the green color was always associated with waters known to contain 
decomposing vegetable matter. In one instance the bilge-water of a sugar ship 
from Cardenas, Cuba, had given such an evolution of this coloring agent that 
Wanklyn’s process was wholly inapplicable as an analytical method on account 
of the impossibility of effecting color comparisons. Experiments with ethylic 
ether and alcohol, with cane sugar, starch, glucose, tannin, and fermenting veg- 
etable matter alwaj s developed the greenish coloration and haziness. More- over, it was noticed that watery dilutions of these substances invariably struck 
a deep yellow color with sodium carbonate; and another reference to former 
laboratory work disclosed the fact that in all cases characterized by the green 
coloration of the Nesslerized distillate, the water had developed a yellow color 
on the addition of sodium carbonate for the liberation of the free ammonia. 
The yellow color thus assumed and the green coloration subsequently interfer- 
ing with the Nessler reaction were therefore regarded as indicating the presence 
of non-nitrogenous vegetable substances in the progress of fermentative 
change. Hence, to summarize: 
A water yielding up the nitrogen of its organic constituents slowly as 
albuminoid ammonia contains recent organic matter,— 
Of animal derivation, if a small quantity of oxygen be required to oxidize it 
by the Kubel or Tidy process,— 
Of vegetable derivation, if a large quantity of oxygen be required. 
A water yielding up the nitrogen of its organic constituents more rapidly 
contains decomposing organic matter,— 
Of animal derivation, if a small quantity of oxygen be required to oxidize 
it, and if there be no interference with the development of the true ammonia 
coloration during Nesslerization,— 
Of vegetable derivation, if a large quantity of oxygen be required, and if a 
yellow coloration be developed in the water on the addition of sodium carbonate 
and a greenish color interfere with the estimation, particularly of the free am- 
monia, by Nessler’s method.