ON ARSENIC 113 



exposed to a gentle heat, and a part of the volatile alkali 

 flying off, it afterwards shows a superabundance of acid. 



(b) When it is thus supersaturated with acid, it forms 

 long radiated acid crystals, but deliquesces again in the air. 



(c) Arsenical sal ammoniac, when distilled, yields a liquid 

 volatile alkali ; it then flows, and arsenic is sublimed ; a 

 white mass remains, which likewise melts in a stronger 

 heat, and proves to be the acid of arsenic. Since the acid 

 of arsenic cannot be converted into arsenic without phlo- 

 giston, and such a change during the distillation cannot be 

 ascribed to the heat, because the arsenic is sublimed before 

 the residuum or acid begins to flow, it follows that the acid 

 of arsenic must decompound part of the volatile alkali 

 during the process. In my experiments on manganese, I 

 observed that a peculiar kind of air is obtained whenever 

 volatile alkali is decompounded. (See Diss. on Manganese, 

 Sec. xxxix.). The same thing happens here : I distilled 

 1 oz. of arsenical sal ammoniac in a small retort, and applied 

 a dry bladder instead of a receiver. Immediately, as the 

 arsenic rose into the neck of the retort, the bladder was 

 filled. This air is perfectly the same with that which is 

 obtained from manganese and sal ammoniac. 



SECTION IX. VITKIOLATED VEGETABLE AND FOSSIL ALKALI. 



(a) I dissolved one part of finely powdered vitriolated 

 vegetable alkali by boiling it in a retort in three parts of 

 acid of arsenic ; the liquid was first distilled off, then a 

 dry receiver was applied, and the fire increased. When 

 the retort grew red hot, the mass melted into a transparent 

 liquor, but no acid went over into the receiver; however, 

 on increasing the fire still further, till the retort began to 

 melt, the mass began to boil, and concentrated acid of vitriol 



