ON ARSENIC 127 



residuum (a) must likewise be reducible by a sufficient 

 degree of heat alone. One part of quicksilver was mixed 

 with six parts of acid of arsenic, and after all the humidity 

 was driven off, the fire was increased, and though the mass 

 was heated white, it, however, lay unchanged in the retort ; 

 but when the retort melted, it ran likewise into fusion ; the 

 quicksilver was reduced, and rose, together with the arsenic, 

 into the neck of the retort, a little quicksilver remaining 

 behind. (d) It is known that quicksilver digested with 

 muriatic acid is not attacked. But if two parts of muriatic 

 acid, two parts of acid of arsenic, and one part of quicksilver, 

 be exposed to digestion in a close phial for fourteen days, 

 the acid acquires the taste of quicksilver, and that metal is 

 changed into a. yellow pow r der. If the acid be now distilled, 

 muriatic acid impregnated with arsenic goes over. If the 

 residuum in the retort be urged by a strong heat, corrosive 

 sublimate rises, and the superabundant acid of arsenic 

 remains behind in the retort. The yellow powder is 

 turned black by lime-water and alkalies ; if sublimed, it is 

 changed into mercurius dulcis. (e) The solutions of quick- 

 silver in vitriolic and nitrous acids are precipitated by the 

 acid of arsenic in the form of a yellow powder ; but the 

 solution of corrosive sublimate is neither precipitated by 

 the acid of arsenic, nor by any of the neutral arsenical 

 salts ; but arsenicated volatile alkali (Sec. vin.) precipitates 

 the solution in the form of a white powder. (/) The acid 

 of arsenic, when sublimed with corrosive sublimate, under- 

 goes no change. But if, instead of corrosive sublimate, 

 mercurius dulcis is employed, corrosive sublimate rises, and 

 the residuum is the same as that of letter (a). Some writers 

 assert that arsenic distilled with corrosive sublimate yields 

 butter of arsenic ; but they are mistaken. I have distilled 

 such a mixture in different proportions, but have always 



