34 PROCEEDINGS OF THE AMERICAN ACADEMY 



0-7938 gr. gave 0-4735 gr. Hg2Cl2 = 50*65 per cent mercury. 

 0-3970 gr. gave 0-3771 gr. silver =31-23 per cent chlorine. 

 0-9752 gr. gave 0-9356 gr. silver =31-42 per cent chlorine. 

 1-3600 gr. gave 0-1024 gr. cobalt = 7*52 per cent cobalt. 



On Blomstrand's view the formulas of the two mercui-y salts may be 

 written 





2 



NH3— a=HgCi 



NH3 — NH3— Cl = HgCl2 

 NH3-NH3-Cl = HgCl, ^ \ 

 NH3 — NH3— Cl = HgC]2 '"''•' ^ 

 NH3 — NH3 — CI = HgCla 

 NH3— Cl = HgCl2 



NHg— CI 



NH3 — NH3 — CI = HgClj 

 NH3 — NH3 — CI = FlgCla 

 NH3 — NHg — CI = HgClj 

 NH3 — NH3 — CI = HgClj 

 NH3— CI 



It is possible that the salt containing six atoms of mercury belongs 

 to the roseocobalt series, as I find that it is formed when a solution 

 of HgClgNa is added to one of the soluble sulphate of roseocobalt, 

 Co2(NH8)io(S04)3 4-5Aq, which I shall describe farther on. I may 

 also remark that if the progress of science should make us acquainted 

 with a method of determining cobalt in these salts with precision, they 

 would enable us to determine the atomic weight of that metal with 

 great accuracy, the first salt containing 5-54, and the second 7-44 

 per cent of cobalt, so that a relatively large error in the weight of 

 (NH3)ioHggClj8, or of (NH3)joHg^Cli4, would exert an inappreciable 

 influence on the result. If we regard the salt Co2(NH8)ioCle -{- 4HgCl2 

 as unsaturated, it ought to combine with other electro-negative chlorides 

 to form salts with three metallic elements. I have not, however, found 

 this to be the case, so far, at least, as the chlorides of gold and platinum 

 are concerned. 



In analyzing these salts, I found it most advantageous to determine 

 the mei-cury in the form of calomel, by dissolving the salt in water, 

 adding a little chlorhydric acid, and then reducing the mercuric to 

 mercurous chloride by a solution of sodic hypophosphite, the solution 

 of mercuric salt having the temperature of 40° C. The mercurous 

 chloride was then weighed on a porous earthenware cone at 100° C. 

 In determining chlorine it is best to dissolve the salt in hot water, with 

 a little free sulphuric acid. The mercury may then be separated as 



