438 Proressor Kane on the Action of Ammonia 
Mercury 86,80 
Chlorine 8,03 
Ammonia 3,87 100,00 
Oxygen and loss 1,30 
In the processes, A and B, a small quantity of the yellow powder was lost, in con- 
sequence of its being not perfectly insoluble in water. This quantity, I have reason 
to believe, varied from one to two per cent. and by the means of calculation we em- 
ployed, the mercury and chlorine constituents are given above what is correct in that 
proportion. As far as I can judge, by considering the circumstances of the ex- 
periments, I conceive the mean to be consequently too high; and I believe the 
analysis C. by itself, to approach closer to the truth. By it we have in 100 of the 
yellow powder— 
Mercury 86,23) 
Chlorine 777 
Ammonia 3,83 100,00 
Oxygen and loss 2,17 
This yellow powder is generated evidently by the reaction of water on white pre- 
cipitate, in which one-half of the chlorine and ammonia are converted into sal- 
ammoniac, a corresponding portion of the mercury being oxydized. I shall compare 
the results of this reaction with each of the formule for white precipitate that I have 
previously given. 
24 (ech + Hg) + (2 Nur + fie) t 4 2H = 
4 (2Ch + Hg) + (8 He + ont) t + 2ChNH! 
Here two atoms of white precipitate and two of water, mutually reacting, give two 
atoms of sal-ammoniac and one of the powder. On this idea, the yellow powder 
should consist of 
Mercury 84,1 ‘| 
Chlorine 7,36 
Ammonia 3,56 100,00 
Oxygen 4,96 J 
These numbers, with the exception of oxygen, fall below the lowest experimental 
result, and probabilities are therefore rather against this formula being true. Let us 
next try the result of reaction with that formula for white precipitate which does not 
include oxygen, and supposes the ammonia to exist as amidogene. 
