f 00 Experime7if.s for ascertaining the Existence of 



to an almost red heat. A small quantity of saline matter,- 

 amounting to about three or four grains, remained, which 

 according to the different proofs made with it was nitrate of 

 anuTfonia, and which when dissolved exhibited the same 

 phenomena as those to be mentioned hereafter in the se- 

 venth experiment. This no doubt arose from a still re- 

 maining trace of the arsenic acid, but by no means froi« 

 cobaltic acid. 



Bxperiinent VI. 



The other tiaTf of the solution, obtained by the fourth 

 experiment, was decomposed by pure carbonate of anuno- 

 nia, by which means a pale red precipitate inclining to vi- 

 olet was produced. The precipitate being then edulco- 

 rated with a sufficient quantity of water, it was digested 

 with twice as much carbonate of ammonia and half an 

 ounce of pure ammonia, by which means a beautiful violet 

 blue solution inclining a little to red was produced, and- 

 there remained 30 gJams of oxide of cobalt undissolved in 

 the state of a carbonate. The filtered solution was evapo- 

 rated in a moderate heat to dryness, by which the ammo- 

 nia was dissipated, and the oxide of cobalt was separated of 

 a dirty ycUov.'ish green colour. When edulcorated and 

 dried, it amounted to 24 grains, and contained less car- 

 bonate than the almost imperceptible residuum of the di- 

 gestion with ammonia. The liquor obtained from the edul- 

 coration of the greenish yellow oxide of cobalt produced 

 by the evaporation of the amnioniacal solution, was evapo- 

 rated in a porcelain dish, and exposed to a strong heat, but 

 without bringing it to a red heat. By these means a resi- 

 duum of saline matter weighing scarcely two grains, and 

 »ome black oxide of cobalt, were obtained. Th? solution 

 prepared from this substance exhibited the same phaeno- 

 Hiena as nitrate of ammonia, and even with reagents, as 

 will be further mentioned in the seventh experiment ; only 

 it appeared to form with nitrate of silver a somewhat 

 stronger precipitate, which could not be again immediately 

 dissolved, and which probably arose from the muriatic acid 

 of the ammonia employed, as there was reason to suspect 

 from its while colour. From these phsenomena it is proved 

 that some traces of arsenic acid were found in the precipi- 

 tate of cobalt' effected by ammonia, but noue of cobaJliq 

 acid.r 



It still remained to examine the liquor in which the ox- 

 ide of cobalt had been separated from its uuloa with nitrous 



aci4 



