204 Proceedings of Royal Society of Edinburgh, [april 18, 
of the cobalt from the cobaltous to the cobaltic state. This seemed 
to point to the probable formation of an alum. Experiments were 
therefore made in order to obtain crystals. The action was allowed 
to go on for several days, so as to concentrate the liquid * when, in 
the solution with ammonium sulphate, dark blue crystals separated 
out. These were octahedral, showing also faces of the cube and 
rhombic dodecahedron. The analysis of these showed that they 
were ammonium-cobalt alum. 
C0 2 0 3 
17-87 
Theory. 
17-05 
(NH 4 ) 2 0 
5-18 
5-36 
GO 
O 
CO 
32-81 
33-02 
H 2 0 (By Difif.) 
44-14 
44-57 
100-00 
100-00 
When thoroughly dried the crystals are not unstable, but when 
dissolved in water are quickly reduced. The solution in sulphuric 
acid is not so soon decomposed. When the acid has been com- 
pletely removed from the crystals, water seems to resolve them first 
into the constituent sulphates. The oxygen liberated during decom- 
position is partially in the form of ozone. 
No satisfactory result was at first obtained with the potassium 
sulphate solution, chiefly owing to the slight solubility of that salt. 
If sufficient be not added the alum is decomposed. If, on the 
other hand, too much be added, it crystallises out. The happy 
medium can only be obtained by the addition of one or other 
salt. After several experiments, crystals were obtained similar to 
those of the ammonium alum, but they could not be separated from 
the potassium sulphate with which they were mixed. Afterwards 
a finely crystalline deposit was obtained, which on analysis appeared 
to be the alum mixed with potassic sulphate. 
Under similar conditions, nickel sulphate undergoes no change 
whatever. 
Other salts of cobalt undergo oxidation in presence of correspond- 
ing alkali salts, forming evidently double salts ; these are at present 
undergoing examination. 
