Cobaltous into Cobaltic Xitrite. g 



suspended until dissolution of the latter was nearly complete. The 

 filtered solution Avas evaporated under reduced pressure at 30°, and 

 yielded the expected crystals which were analysed. The results 

 agreed well with Rosenheim and Koppel' s, except that the latter 

 did not include any finding for the cobaltic element, the propor- 

 tion of it present being assumed by them to be two-fifths of the 

 total cobalt. Calculation for Co"4Co"'2(NO,)«(N03),(OH)4,15H20 

 gives figures which, with those here obtained and those found by 

 Rosenheim and Koppel, are as follows: — total cobalt, f. 30-27, 

 c. 29-90, Ros. f. 30-15; cobaltic cobalt, f. 9-80, c. 9-97; total 

 nitrogen, f. 11-77, c. 11-82, Ros. f. 11-05; nitrite nitrogen, f. 9-31, 

 c. 9-46, Ros. f. 8-90. Rosenheim and Koppel gave the salt a formula 

 equivalent to Co';Co";(N02),(OH)4,Co(N03)2,12H20; but this 

 can be set aside, now that the ratio, Co":Co"' has been experi- 

 mentally determined. In the hght thrown upon the constitution 

 of the hydroxycobaltinitrites by the work described in the 

 present paper, the constitution of this nitrate- containing cobaltous 

 salt may be expressed by 



N03Co"NO., N0.,\ /OH 



NO.,Co"NO., NO.,— Co'".,— OH OHCo"NO.„ 15H.,0, 



NO,Co"NO, NO,/ \0H 



which differs from the formula given on p. 3 only in having 

 2NO3 in place of 2NO2. But the crystals obtained from the solution 

 prepared by Lang's method were more basic than this, and, 

 in accordance with this difference, it was found that the 

 crystals obtained from the solution prepared by Rosenheim and 

 Koppel' s method do not, when treated with solution of sodium 

 carbonate equivalent to half their cobaltous element, give a 

 precipitate containing noticeably more than half the cobaltous 

 element. In the case of the crystals from the Lang solution, the 



