644 
Proceedings of the Royal Society 
permanganate, and therefore of oxygen required, is to follow such a 
process as Feldhaus describes in a later paper* than that I have 
referred to. He recommends the nitrite to he dissolved in very 
slightly acidulated water, and permanganate of potassium to he 
added until the oxidation of the nitrous acid is nearly completed, 
when the solution is to he made strongly acid, and permanganate 
run in until a permanent light red coloration is obtained. How, 
as Fresenius f points out, and as very little experience teaches us, 
nitrous acid, owing to its great tendency to decompose in watery 
solution with the formation of nitric acid and nitric oxide, requires 
in such an estimation as that we are considering to he present in the 
amount of not more than 1 in 5000 of water, and even with this, 
and a much greater degree of dilution, decomposition slowly pro- 
ceeds. Furthermore, we must remember anhydrous nitrous acid 
(H 2 0 3 ) — we cannot speak of the hydrated acid (HH0 2 ), as it is not 
known in the pure state — has a boiling point of about - 10° C., J so 
that the risk is incurred of the estimation being vitiated not only 
by decomposition, hut by decomposition and volatilisation combined. 
After numerous experiments with this method of Feldhaus, I have 
not obtained satisfactory results. The slowness of oxidation which 
occurs when the feebly acidulated solution contains hut little per- 
manganate, gives time for those evils — volatilisation and decomposi- 
tion— of which I have just spoken, to occur. Moreover, since there 
is some difficulty in knowing when the reaction is almost, and even 
more when it is quite, completed, I have been led to look for a 
modification of the permanganate process — a modification which will 
allow very rapid oxidation of the nitrous to nitric acid. To add 
some excess of permanganate, and estimate that excess by a reducing 
agent, has been frequently described, hut I obtain the most satisfac- 
tory results by using a large excess of permanganate, and titrating 
hack with some suitable reducing agent. Such an agent may he 
found in oxalic acid, as suggested by Kinnicult and Hef,§ or in 
ammonio-ferrous sulphate, as employed by Feldhausen and Kubel.|| 
* Zeitschrift fur analytische Chemie, vol. i. p. 426. 
t Anleitung zur quantitative chemischen Analyse , 6th edit., vol. i. p.. 
390. 
t Dittmar, Chemical Analysis, p. 224. 
§ American Chemical Journal, vol. v. p. 388. 
|| Classen, Quantitative Analyse, p. 222. 
