356 PROCEEDINGS OF THE AMERICAN ACADEMY 



formed and analyzed the compound with perchloride of iron. Its 

 formula is 



Fe2 CI3, 4 HgCy, 7 HO. 



I hope to get similar salts with AljCls, BegClsjand perhaps with 

 Cr. CI3. 



For the salt of cyanide of mercury and chromate of potash, first 

 described by Caillot and Podevin,* Rammelsberg f has found the 

 tbrraula 



2 (Ko,Cr03), 3 HgCy, 



which was changed by Poggiale + to 



(Ko,Cr03), 2 HgCy. 



An analysis of this salt has given me results agreeing very nearly 

 with those of Rammelsberg, with the addition of one equivalent 

 of water, which has been hitherto overlooked. Its formula would 

 then be 



2 (Ko,Cr03), 3 HgCy, HO, 



The analyses of the compound which has been mentioned as con- 

 sisting of 



BaCl, 2 HgCy, 6 HO 



have shown that the composition of this salt is not constant, and is not 

 in exact accordance with the laivs of chemical proportion. The above 

 formula requires 16.73 Ba, and 48.77 Ilg in the hundred. In the salt 

 as I have obtained it, the barium is always deficient in quantity and 

 the mercury in excess. The barium has been found as low as 13.4, 

 and I have never found it higher than 15.G9 ; wbile the mercury 

 varied from 54.3 to 50.5. In general, the smaller the excess of chlo- 

 ride of barium in the solution from which it crystallizes, the less ba- 

 rium and the more mercury will be found in the salt. In some of the 

 cases, those which gave the extreme numbers, this may very probably 

 be owing to a mechanical admixture of cyanide of mercury, the crys- 

 tals of which formed at the same time with those of the double salt, 

 and, as I shall on another occasion show, cannot always be distin- 

 guished from them. It is possible, too, that there may be a compound 



* Berzelius, Jahrsb., VI. 183. 



t Pogg. Ann., Bd. XLII. S. 131, and Bd. LXXXV. S. 145. J Loc. at. 



