CHEMISTRY OF NITROGLYCERINE. 79 



used, or a proportion exactly similar to that produced by the action of a fixed 

 alkali.'"" 



Action of Alkaline Carbonates. — Nitroglycerine was dissolved in slightly 

 diluted alcohol in order to permit the solution in it of carbonate of potash, 

 which was afterwards added in the form of a concentrated aqueous solution. 

 The mixture slowly assumed a reddish colour even in the cold. Heated over 

 the water-bath, the nitroglycerine tolerably rapidly underwent decomposition, 

 and the fluid became of a fairly deep reddish-brown colour. The. fluid was 

 heated for one hour, and water was added as the alcohol became evaporated. 

 The yield of nitrous anhydride was 35*24 per cent., or the same as when 

 caustic potash is used. In a second experiment the yield was exactly the 

 same. Using a five per cent, alcoholic solution of nitroglycerine, and boiling 

 it with excess of carbonate of potash, it was ascertained that complete 

 decomposition occurs in about ten minutes. 



Action of Phosphate of Soda fNa. 2 HPO J. — This salt is acid in constitution, 

 but alkaline in reaction. Added in concentrated aqueous solution to a one 

 per cent, alcoholic solution of nitroglycerine, and heated over the water-bath, the 

 phosphate commenced to decompose the nitroglycerine three to four minutes 

 after the mixture was fully heated, as was evidenced by the appearance of a 

 yellowish tint gradually deepening to an orange-red. After heating for an 

 hour and a half, slightly diluted alcohol being added occasionally, a yield of 

 1376 per cent, of nitrous anhydride was obtained. In another experiment, 

 after heating a similar mixture for forty minutes, 13*21 per cent, of nitrous 

 anhydride was obtained. On both occasions the nitrous acid was estimated 

 in one half of the fluid, and it was observed, in diluting the fluid with 

 water, that a considerable amount of nitroglycerine was precipitated. In 

 the second experiment, by adding potash to the remaining half of the fluid, 

 and heating for a few minutes, and then estimating the nitrous acid, 34*41 per 

 cent, of nitrous anhydride was obtained, proving that no nitrous acid had been 

 set free by the phosphate of soda, which, in the absence of sufficient alkali, 



* It is somewhat remarkable that the yield of nitrous acid was as g^eat as when a fixed alkali 

 was used ; since it is well known that a solution of nitrite of ammonia readily undergoes decom- 

 position when heated ; and it was to be expected that the estimated yield of nitrous acid would have 

 been diminished by the occurrence of such a decomposition in the heated fluid. In order to make 

 certain that the result obtained was correct, and that under the conditions of the experiment such a 

 decomposition did not occur, or occurred only to a limited extent, I have, since the paper was read, 

 repeated the experiment. This I have done on two occasions, and with separate quantities of pure 

 nitroglycerine. On the first occasion only 2 9 "6 per cent, of nitrous anhydride Avas obtained, but, on 

 the second, 34.1 per cent., or very nearly what was obtained inthe experiment recorded in the text. 

 The nitrite of ammonia formed by the splitting up of the nitroglycerine would appear, therefore, to 

 undergo very little decomposition under the conditions in which it is placed. This may be due to 

 the excess of free ammonia always maintained in the decomposing mixture, or to the low degree of 

 heat applied, as, in order to prevent the rapid escape of gaseous ammonia, heat was applied short of 

 ebullition of the fluid, or, more properly, of the escape of ammonia in the form of bubbles of gas ; or 

 •it may be that the highly concentrated alcohol in which the nitroglycerine was dissolved hindered the 

 decomposition of the nitrite. 



