232 Proceedings of the Royal Society 
author either that nitroglycerine was not a nitrate of glyceryl as 
commonly represented, but contained nitrous acid in its constitution, 
or that it was readily decomposed in the body with the formation of 
nitrous acid. This suggestion led to an extended investigation of 
the constitution of nitroglycerine and of the decomposing action of 
various substances on nitroglycerine. The conclusion arrived at as 
to the constitution of nitroglycerine was confirmatory of the com- 
mon belief that it is a tri-nitrate of glyceryl. For it was found that 
although nitroglycerine yielded a large quantity of nitrous acid 
when decomposed by alkalies, yet nitroglycerine could not be 
obtained by acting on glycerine with nitrous anhydride gas, nor did 
the presence of urea prevent its formation when it was prepared in 
the usual way by allowing nitric and sulphuric acids to act on 
glycerine. The yield of nitroglycerine from a given weight of 
glycerine also closely agreed with its being a tri-nitrate, and was 
considerably in excess of the yield obtainable had nitrous acid 
entered into its constitution. An analysis of the elementary com- 
position of nitroglycerine, which was carried out in conjunction 
with Mr Masson,* also led to the same opinion. 
With respect to the decomposing action of various substances on 
nitroglycerine, the novel and interesting result was obtained that 
potash, soda, ammonia, or an alkaline carbonate decomposed nitro- 
glycerine, when heated with it for a longer or shorter period, form- 
ing an amount of nitrous acid corresponding almost precisely to the 
reduction of two-thirds of the nitric acid which nitroglycerine 
contains. It was also found that the proportion of alkali necessary 
for this decomposition was 5 molecules of the alkali to 3 
molecules of nitroglycerine, and, therefore, a much larger proportion 
of alkali than has hitherto been believed to be necessary. Further, 
in direct contradiction of all previous statements, no glycerine could 
be discovered amongst the products of decomposition, which, besides 
nitrite of the alkali, consisted of nitrate, acetate, formate, and 
oxalate of the alkali, a reddish aldehydic resin, and a body capable 
of gelatinising in alcohol. The amount of nitrous acid was not 
atfccted whether alcohol or water was used as the medium in which 
the alkali was allowed to act on the nitroglycerine. 
A solution of phosphate of soda (NagHPOI) acted on nitro- 
* Vide following paper. 
