10 Art. 2.— T. SiTZuki; The Change of 



precipitate, similarly obtained, did include much more than this, 

 though not as much as the formula indicated might have been 

 precipitated. In that case some, in this case all, the cobaltous 

 hydroxide remained in solution as a more basic hydroxy- 

 cobaltinitrite (p. 4). 



From the Rosenheim and Koppel solution the same potassium 

 hydroxycobaltous salt and the same silver salt were obtained as 

 from the Lang solution. In the statement on p. 7 the second 

 analj^sis of the silver salt was made on a precipitate from such a 

 solution. Rosenheim and Koppel failed to get from it anything 

 but silver nitrite in this way, though they got a different, greyish- 

 brown, microcrystalline, hydrated silver h3'droxycobaltinitrite, 

 indirectly through a zinc hydroxycobaltinitrite prepared in their 

 way. To their silver salt and the other hydroxycobaltinitrites 

 described by them, they have given formulje equivalent to 



Ag,Co"',(NO,)G(0H), + H2O, 

 Ba2Co";(NO,)8(OH). + 0-2Co(NO3)2 + llH,,O, 

 Sr,Co";(NO,),(OHX + lOH.O, 

 Zii,Co'",(NO,)„(OH), + 9K,0. 



But more evidence is wanted to carry conviction as to the 

 existence of such differences as those presented between the 

 formul£G of the strontium and zinc salts and between the composi- 

 tion of Rosenheim and Koppel' s series of salts and that of the 

 salts now described. There is no evidence recorded in their paper 

 as to whether the presence of additional nitrogen as nitrate was 

 tested for in any other case than that of the barium salt, although 

 in all cases the mother-liquor of the salt contained nitrate. Also 

 evidence is wanting of the proportion of cobaltic element in the 

 salts. As a consequence of the latter omission in the paper and b}' 

 taking the more probable view that the nitrate in the barium salt 



