316 VARRENTRAPP AND WILL'S METHOD OF ANALYSIS. 
ART. LIV.— RESEARCHES ON MM. VARRENTRAPP AND 
WILL'S METHOD OF ANALYSIS. By M. Reizet. 
M. Reizet has given an attentive examination to the new 
process recommended by MM. Varrentrapp and Will, for 
the determination of Nitrogen in the analysis of organic sub- 
stances. This process* is founded on the general law, that 
organic substances are decomposed under the influence of the 
fixed alkalies, into water, and carbonic acid, and, when nitro- 
gen is present, ammonia. It results from the experiments of 
M. Reizet, that this process is not free from sources of error. 
In the first place, the mixture of lime and soda retains a por- 
tion of atmospheric air in a peculiar state of condensation; 
this air cannot be separated either by a current of gas or by 
the influence of a vacuum. During combustion the nitrogen 
of this air gives rise to ammonia, and consequently increases 
the resulting amounts of this compound. Faraday has some 
time since remarked that, organic bodies, not containing ni- 
trogen, carbon itself, and those metals which decompose wa- 
ter, yield ammonia, when calcined in contact with air and 
potassa. 
Another source of eiror is, that the alcohol in which the 
perchloride of platinum is dissolved, reduces this salt into in- 
soluble protochloride ; this reduction takes place slowly — 
nevertheless there is sufficient of the protochloride formed 
to increase the weight of the ammoniacal salt of platinum, and 
consequently augment sensibly the estimation of the nitrogen. 
It is inexplicable how MM. Varrentrapp and Will have al- 
ways obtained less nitrogen than theory would indicate in the 
objects of their analysis, since the cause of error would tend 
to give an excess, unless it be admitted that, during the pro- 
* See page 141 of this Volume. 
