74 DR MATTHEW HAY ON THE 



four-fifths of the fluid had been evaporated. An equal bulk of water was now 

 added to the residue, and distillation was continued until the alcohol was almost 

 completely expelled. The residue was now saturated with pure barium hydrate, 

 in order to remove the sulphuric acid, and then filtered ; and the excess of 

 barium was precipitated by a stream of carbonic acid gas, and the fluid was 

 agaiu filtered. The filtrate was now evaporated over the water-bath, and was 

 quickly reduced to five or six drops of a golden yellow viscid residue. Treated 

 with absolute alcohol, in which glycerine is freely soluble, it at once hardened, 

 and was almost entirely insoluble in alcohol. It evidently consisted in part of 

 a barium salt, as ascertained by testing. The alcoholic solution or extract was 

 filtered, and, on evaporation, yielded about one drop of a yellowish syrup, much 

 more viscid than glycerine, and pungent rather than sweet to the taste. A few 

 minutes' further drying dessicated it to a hard scale. The syrup gave merely 

 the faintest odour of acrolein when heated with acid sulphate of potash. J 

 therefore concluded that this residue, which ought to have contained the greater 

 part of the glycerine, were any present, contained practically none of that sub- 

 stance. The absence of glycerine from the alkaline decomposition products of 

 nitroglycerine was confirmed by a second experiment, made with a still larger 

 quantity of nitroglycerine, and in which no distillation was practised, and less 

 opportunity therefore afforded for the decomposition or evaporation of the 

 glycerine. 



The deep reddish-brown residue, laid aside at the commencement, after 

 being well washed and extracted with boiling absolute alcohol, was next 

 examined. Dried at 100° C. for 24 hours, it weighed 14*65 grms., and was 

 probably not even then absolutely dry,, although very nearly so. The large 

 amount of the residue is remarkable, as it weighs considerably more than the 

 sum of the weights of the nitroglycerine decomposed, and of the potash neces- 

 sary, according to Railton's equation, for the decomposition of the nitro- 

 glycerine. This point will shortly receive an explanation. 



The dried residue was perfectly soluble in water, forming even with a large 

 volume of water a deep reddish-brown solution. A portion of the residue was 

 dissolved in boiling water, and, after standing for some hours, large needle- 

 shaped crystals of nitrate of potassium separated, which by re-solution and 

 re-crystallisation were obtained in a perfectly pure form, and distinctly recognised 

 to be nitrate of potassium. From another measured portion of the residue it 

 was attempted to remove the nitrite of potassium and formate of potassium (if 

 present), by treatment with strong acetic acid, and extraction with boiling 

 absolute alcohol, in order to remove the acetate of potash and free acids thus 

 formed, and to obtain in a tolerably pure state the nitrate of potassium for the 

 purpose of a quantitative determination ; but the method did not succeed. 



One-half of the original dried residue was next dissolved in water, and 



