6 PKOCEEDINGS OP THE AMERICAN ACADEMY 



0-4392 gr. gave 0-2332 gr. SO,Co = 20-21% cobalt. 

 0-7952 gr. gave 0.9208 gr. SO,Ba= 47-71% SO^. 



These analyses lead to the formula : — 



Co,(NH3),„(SO,)3+OH,. 



Calculated. Found. 



Cobalt, 2 20-27 20-21 



SO,, 3 47-42 47-71 



The results of the analyses agree well with the formula given ; but it 

 is possible that the salt, which was not recrystallized, contained some 

 small impurity, and that it is really anhydrous, like the normal chloride 

 and sulphate. The sulphate is but slightly soluble in cold water, but 

 di.-solves rather easily in hot water, forming a beautiful violet solution. 

 Auro-chloride of sodium gives a beautiful crystalline orange-red pre- 

 cipitate with a solution of the sulphate, differing apparently from the 

 corresponding salt of roseocobalt. In this salt, 



0-2698 gr. gave 0-0924 gr. gold = 34-24%. 



This percentage of gold corresponds ver}' closely to that required by 

 the formula, Co,(XII,)i,j(SO,).,Cl,4-2AuCl„ which is 34-02. The 

 sulphato-chloro-aurate of roseocobalt has, as I shall show, the formula 

 Co^,(XH.j),„(SO,)2Cl2-|-2AuCl3-)-40H2, the purpureo-salt showing, as 

 usual, a less disposition to unite with water of crystallization. The 

 above results were obtained with only three grammes of material ; but 

 they are sufficient, I think, to make it at least probable that the violet 

 salt in cpiestion is the true sulphate of purpureocobalt. 



Pyropliosplidte. — Sodic pyrophosphate gives a lilac or rose colored 

 preci|)itate with a solution of chloride or nitrate of purpureocobalt, 

 readily soluble in an excess of the precipitant, and crystallizing from 

 the solution in beautiful rose-red efflorescent crystalline scales. The 

 salt is readily soluble in ammonia, and the solution yields beautiful 

 garnet-red measurable crystals, which are free from sodium. The pyro- 

 j)ln)spliate was first carefully studied and analyzeil by C D. Braun, 

 whose analyses agree closely with a formula which he writes, 



oNii3.CoA.2PO,,4-2i no 



(old style). This formula must now be written 

 Co,(NH3),„(r,0,,)-h2lOH, 



in the new notation, P,0i3, being hexatomic. In the salt crystallized 

 from ammonia : — 



