212 
NATURE 
| Dec. 28, 1882 
Special conclusions as to the Albuminoid-ammonia Process 
1. In the determination of both ‘‘free” and ‘‘albuminoid” 
ammonia there is a loss, which may be quite considerable, from 
imperfect condensation of the ammonia during distillation, This 
varies especially with the efficiency of the cooling apparatus and 
the time occupied. 
2. Where waters contain urea and other amidated bodies, 
such as the leucine and tyrosine of putrefactive decay, some 
ammonia is so easily formed from these substances by boiling 
with sodium carbonate, or even without this addition, that it is 
impossible to distinguish sharply between pre-existing ‘‘ free” 
ammonia (of ammoniacal salts), and that formed by the action 
of alkaline permanganate, the so-called ‘‘albuminoid ” ammonia. 
This sourcefof error as to free ammonia reacts (as noticed above), 
on the result for organic nitrogen by the combustion process. 
3. There is no satisfactory evidence for Wanklyn’s view, that, 
in distilling with alkaline permanganate, definite and simple 
fractions of the nitrogen of organic matter are given off as 
‘‘albuminoid” ammonia. Such results may be varied at plea- 
sure for most substances, by modifying the conditions of dis- 
tillation. 
4. If the distillation with alkaline permanganate be carried 
out according to Wanklyn, the nitrogenous organic matter is 
often so gradually acted upon, as to make the ending of the 
process indefinite ; ammonia is sti]l coming off when the distilla- 
tion has to be stopped, the contents of the retort being nearly 
dry. Here an unknown fraction of the possible amount of 
albuminoid ammonia fails to be collected, and is but vaguely 
indicated by + after the figures recorded. 
5. There is evidence that in some cases nitrogenous matter is 
volatilised during the distillation for free ammonia, which, if it 
had been retained, would have yielded up its nitrogen as albu- 
minoid ammonia. Not affecting the Nessler reagent, such 
matter escapes detection. 
6.-The albuminoid-ammonia process proper, z.e. distillation 
with alkaline permanganate and determination of the ammonia 
evolved, is admittedly simple, and easily carried out with very 
little preparatory training. 
7. The value of the resu’ts by this process depends more on 
watching the Avogress and rate of evolution of the ammonia than 
upon determining its total amount. 
8. The recorded results by this process show a good deal of 
similarity between the figures for albuminoid ammonia and those 
for organic nitrogen (by the combustion process), but with fre- 
quent discrepancies of varying extent, such as prevent the one 
being taken as the accurate measure of the other. 
Special Conclusions as to the Permanganate Process 
1. The results by the Tidy method (the acidified permanganate 
at coulmon temperature) and that of Kubel (operating at boiling 
point) differ irregularly from each other, the latter usually giving 
much higher figures, as was to be expected, but the ratio between 
the results by the two methods varies much in different cases. 
2. On the whole, there seems to be a nearer approach to pro- 
portionality with the quantities of organic carbon found by the 
combustion process on the part of the Kubel process than on 
that of the Tidy process, but to this there are some very notable 
exceptions. 
3. In a good many cases, the Kubel results are, contrary to 
the general rule, lower than those by the Tidy method. Thi 
seems to be due to loss of organic matter by volatilisation with 
the escaping steam from the boiling liquid before time has been 
afforded for action on the permanganate. Of course, a similar 
loss may have occurred in other cases, but not to the extent of 
reversing the general rule; and this may in part explain the 
absence of any uniform ratio between the figures yielded by the 
two methods. 
4. The results by the Tidy process are liable to variation with 
the atmospheric temperature at the time of operation. 
5. The amount of oxygen consumed by a specimen of water 
is probably in all ordinary cases much below that required for 
complete oxidation of the organic matter present, and does not 
stand in any fixed ratio thereto ; it cannot be taken either as a 
measure of the organic carbon or of the total organic matter. 
Still a distinct general resemblance may be traced between 
strongly marked results, high or low as the case may be, for the 
consumption of oxygen on the one hand, and organic carbon (by 
the combustion process) on the other, and closer agreement is 
observable regarding waters of generally similar character. 
6. The permanganate process is capable of giving more valu- 
able information in regard to a water by watching the progress 
and vaée of the oxidation of organic matter present than by any 
single determination of the actual amount of oxygen consumed 
in a given time. 
7. For such observation of the progress of oxidation, the two 
determinations prescribed by Tidy, viz. of oxygen consumed in 
one hour and three hours respectively, are not sufficient, nor is 
the latter period of three hours long enough to indicate the 
general behaviour of the water with the acid permanganate. 
As to other chemical determinations, the discrepancies in 
those of total solids left on evaporation, of the loss on ignition 
of this solid residue, and even of chlorine, are noted as illus- 
trating the comparative roughness of the methods with which 
those results were obtained when very small quantities have to 
be dealt with. A coincidence is often presented between alka- 
line reaction of a water and the occurrence of nitrites and 
nitrates in considerable quantity, suggesting a recollection of the 
conditions under which nitrates are produced on a large scale in 
the decay of nitrogenous organic matter. Those salts, how ever, 
also occur pretty largely in some cases without alkaline reaction, 
and in other cases there was alkaline reaction and also much nitro- 
genous matter, but no nitrites nor nitrates. Ammonium nitrates 
seem to be rare, the basis-constituent being rather mostly non-vola- 
tile—no doubt calcium, magnesium, or one of the alkaline metals ; 
this is noticed as in relation to possible reduction to nitrite, and 
consequent loss of nitrogen in the combustion process. Exyeri- 
ments with tannin showed the utter worthlessness of this gr. up- 
reagent of Kammerer for the purpose for which he has advocated 
its use. The general series of analyses of dissolved gases ii!us- 
trate how the results are influenced by varying condition; of 
oxidisability of organic matter present, temperature, extent of 
exposure to the atmosphere, and interchange with it, in both 
direction, of gaseous constituents. 
With regard to the microscopic and pathological results, Prof. 
Mallet, feeling himself incompetent to properly discuss these, 
prefers leaving them to speak for themselves, and merely 
remarks on some of the difficulties of such research, and, at the 
same time, on its importance and value when rightly conducted. 
Passing now to sanitary conclusions and interpretation of 
results, the Report deals with 
Chemical and Biological Results as contrasted with the actuas 
Sanitary History of the Natural Waters Examined 
Now on inspection of the tabulated results it appears that oe 
strongly marked generic difference is presented by the results from 
any of the processes for estimaticn lof organic matter or tts 
elements between the generally wholesonie waters of Class 1. and 
the waters of Class II, medically condemned and fairly assumed 
as pernicious. ‘This applies equally to the highest, the lowest, 
and the average results. No one could, with those figures to 
guide him, refer a water of unknown origin to one or other of 
the two classes, on the basis of chemical analysis by any or all 
of the three methods, Attention is called to the smallness of 
the amount of organic matter indicated as present in many of the 
most dangerous waters, giving important evidence against any 
chemical theory of the production of di-ease from this source (on 
the simple assumption that some of the chemical products of 
decomposition of organic matter are poisonous or noxious in 
their effect on the human system). Thus in the case of two 
waters of highly dangerous character, if the whole of the organic 
carbon and nitrogen present existed as strychnine, it would be 
necessary to drink about half a gallon of the water at once, in 
order to swallow an average medicinal dose of the alkaloid. It 
is not easy to believe that the ptomaines, or other chemical pro- 
ducts of putrefactive change, can be so much more poisonous 
than the strongest of recognised poisons, While most of the 
mischief in drinking water is probably attributable to living 
organisms, the possibility is noted, that indirectly a large amount 
of-organic matter in water may be more dangerous than a 
smaller quantity, as furnishing on a greater scale the suitable 
material and conditions for development of organisms, Whether 
variations in the mere quantity of organic matter within such 
limits as occur in water likely to be used for drinking are of 
much importance in this respect, is a question on which (in the 
author’s opinion) depends largely the utility of all attempts to 
estimate the quantity of organic matter or its constituents as 
such, 
A much more conspicuous difference between the waters of 
Classes I, and II. is presented by the results for nitrites and 
nitrates. These salts are either absent or present in but trifling 
