376 
ON WATER ANALYSIS. 
BY DR. E. FRANKLAND, F.R.S. 
In pursuing the ordinary processes for the analysis of drinking-waters, the 
only difficulties of any magnitude that present themselves are in the determina¬ 
tion of the organic constituents, and of those substances which result from their 
partial oxidation by exposure to the influence of the air, namely, the ammonia 
and the nitrates and nitrites. Dr. Hofmann, in quitting the professorship at 
the Royal College of Chemistry, left behind him, as a legacy, the monthly exa¬ 
mination of the waters supplied to the Metropolis by the various companies. 
The numerous samples in this way supplied furnished abundant opportunity for 
gaining experience in waters of every degree of purity ; and the analyses neces¬ 
sarily conducted speedily brought about the conclusion that the methods hitherto 
in use for estimating the amount of organic impurity were completely value¬ 
less. 
As has been shown by Messrs. Chapman and Smith,* many forms of organic 
matter urfdergo change during the evaporation of their aqueous solution in such 
a manner that a considerable loss is experienced in that process. For instance, 
urea in one test-experiment lost, when evaporated with carbonate of soda in the 
ordinary way, no less than 44 7 per cent, of its weight. But in addition to 
this and to the loss of ammonia by this the usual method of proceeding, there 
are numerous other sources of error which have generally been overlooked. 
Urea and urates are substances which may, with great probability, find their 
way into well and river water ; and when such substances are evaporated with 
carbonate of soda, and the dry residue is ignited for the purpose of burning off 
volatile and combustible matters, as very often practised, these nitrogenous 
compounds do not suffer complete expulsion,—a result which may probably be 
attributed to the formation of a cyanide or other similar compounds. 
Again, in the employment of carbonic acid water for recarbonating the residue 
after ignition, it frequently happens that the effect of such treatment is posi¬ 
tively to increase the weight of the mass beyond that of the total residue left on 
evaporating the water ; and this may come about, making full allowance for 
any non-volatile matter the carbonic acid solution may itself contain. The in¬ 
fluence of the nitrates contained in the residue is well known; they, of course, 
suffer decomposition upon being heated with carbonaceous substances. 
The employment of permanganate of potash for testing water has been 
already described, and its delusive indications pointed out.f 
With regard to the estimation of the nitrates contained in water, it was 
hoped that Pugh’s extremely elegant process would have answered for that pur¬ 
pose; but Chapman and Schenck have demonstrated that starch, sugar, and 
other organic substances cause a change in protochloride of tin, so as to preclude 
its employment in their presence. Experiments at a much lower temperature 
than that recommended by Pugh were also unsuccessful in establishing its ap¬ 
plicability to the analysis of the residues of waters. 
The ammonia is now generally determined by the use of what is known as 
the “ Nessler testbut it must be observed that, as already noticed elsewhere, 
the test cannot be applied without previous distillation of the water, since many 
forms of organic matter produce a colour upon application of the test, so as to 
interfere with the reaction given by the ammonia. 
The processes now for the first time proposed are comprised under several 
distinct operations. A determination is effected, not of the weight of organic 
matter, but, in the first place, of two of the constituents of the organic matter 
—the carbon and nitrogen; secondly, of the nitrates and nitrites; and lastly of 
* See page 335. 
f See page 294. 
