on the new method for the estimation.of Nitrogen, Sc. 287 
presented an essay to the Academy of Sciences at Paris*, in 
which he endeavours to prove by experiments, at first sight 
very convincing, that the above-mentioned method is attended 
by two sources of error; the first of which in particular, were 
it true, would be quite sufficient to destroy completely the 
value of the method. 
The cause of this first and principal source of error is, ac- 
cording to M. Reiset, that the nitrogen of the atmosphere 
forms a portion of the ammonia produced by the decomposi- 
tion of nitrogenous matter by means of an alkaline hydrate, 
and that consequently too large an amount of nitrogen must 
always be obtained, particularly in bodies rich in carbon, 
bodies of difficult combustion, and those which readily form 
cyanogen compounds. ‘This source of error becomes the 
more important, as from the experiments of Faraday +, which 
are confirmed by Reiset, it appears that by the fusion of many 
non-nitrogenous bodies with the hydrates of the alkalies a 
pretty considerable quantity of ammonia is formed. To those 
non-nitrogenous bodies which produce aminonia belongs in 
particular sugar, a substance which we proposed as an addi- 
tion for the purpose of diminishing the violence of the absorp- 
tion of the ammonia by the hydrochloric acid. 
The numerous analyses of nitrogenous bodies made by 
Varrentrapp and myself, must have given a very considerable 
excess of nitrogen, if any formation of ammonia really took 
place, and were a constant source of error: it would be parti- 
cularly evident in the analyses of ammeline, in which we mixed 
the substance with an equal weight of sugar. The accuracy 
and strictness of the results obtained by us from substances 
of well-known composition could therefore be ascribed only 
to accident, or perhaps to some source of error balancing the 
one just mentioned. 
We thought we had met every objection of a source of 
error on this point by the experiments mentioned in our paper, 
in which we passed nitrogen and hydrogen gases through a 
glass tube over a red-hot mixture of carbonized bitartrate of 
potash and lime, or of pure charcoal soda and lime, and from 
which we did not obtain sufficient ammonia to be estimated 
as ammonio-chloride of platinum; and yet all the conditions 
necessary for the formation of ammonia and cyanogen were 
afforded in the mixture of soda, lime and carbon by the hy- 
* Compt. Rend., vol. xv. p. 154; and Annal. de Chim. et de Physique, 
3rd ser. vol. v. p. 469. 
+ Quarterly Journal of Science, vol, xix. p. 16; and Poggendorff’s dn- 
nalen, yol, iii. p.455, [Also Phil. Mag, S. 1. vol. Ixv. p. 309.) 
