176 TEST FOR ARSENIC. 



part of the tube, in minute octoedral crystals of arsenious 

 acid. 

 It is an arte- It appears, therefore, that the precipitate in question is a 

 nite ol silver, combination of white arsenic (arsenious acid) and m!v< »", or an 

 arsenite of silver j and it is inferred that its formation, when 

 ammonia and nitrate of silver are added to a mixture containing 

 arsenious acid, is owing to a double elective decomposition of 

 the arsenite of ammonia, by the nitrate of silver, in conse- 

 quence of which arsenite of silver is formed, and separates as 

 an insoluble precipitate from the nitrate of ammonia which 



_. . . remains in the solution. The addition of ammonia is neces- 



Vse and ad- ; / c 



vantage of the sar > T » because arsenic acid alone cannot decompose nitrate ot 

 ajimonia. silver ; but in Fowler's solution, in which the arsenic is already 

 combined with an alkali, the decomposition takes place at 

 once, without any addition of ammonia. The fixed alkalies, 

 therefore, can answer a similar purpose ; but ammonia lias this 

 advantage, that it does not, wiien added singly, decompose 

 nitrate of silver, a circumstance which, in using the fixed alka- 

 lies, might occasion &ome confusion*. 

 Mr.SylveMer's With regard to Mr. Sylve.^er's objection, 1 shall, previous 

 that muriatic to m J r offering any remarks upon it, state it in his own words. 

 acid would, if " If ever muriatic acid be present," says this gentleman, 

 theTilvcr! € Z6 '' l ^ e test * s tnen wn °lty useless, as a muriate of silver will be 

 immediately formed, and the yellow compound, said to be so 

 unequivocal in its indication of arsenic, of course be pre- 

 vented from appearing." 



This clanger of ambiguity, however, though applying in 

 some degree to the process in question, and well deserving to 

 be noticed, will be found to have been greatly overrated ; and" 



* It is necessary, as Dr. Roget has observed in the paper already 

 quoted, that the quantity of ammonia should not be too large ; for in 

 that case the precipitate is redissolvcd. But, even then, it may be 

 made to reappear, by the addition of nitric acid in sufficient quan- 

 tity to saturate the alkali. In this case, however, the precipitate is 

 not permanent, owing-) I find, to its being soluble in the nitrate of am- 

 monia which is formed in the process. Carbonate of ammonia has also 

 the propei ty of producing and redissolving the precipitate. 



The fixed alkalies in excess have not the power of redissolving the 

 precipitate. 



there 



