176 Mr Wood and Mr Jones, On Complex Carbonates. 
but we were never able to get analytical results which shewed 
that it was anything but a mixture of varying composition. 
A solution of potassium carbonate containing 100 grams to 
100 grams of water was saturated with Groger’s salt at 25° C., 
the clear dark blue solution was decanted or filtered and on 
allowing it to stand, it set to a mass of fine needles which when 
filtered, dried and analysed were found to be K 2 C0 3 CuC0 3 H 2 0. 
The filtrate was found to contain 3T1 grams copper carbonate 
and 80 grams of potassium carbonate in 100 grams of water, and 
its composition is represented by the point E in the figure. 
In studying the behaviour of other metallic salts towards 
concentrated potassium carbonate solutions, it was observed that 
cobalt salts behave in a remarkable way. When a concentrated 
solution of cobalt chloride is added to potassium carbonate 
solution a precipitate is formed, which dissolves on shaking to 
give an intensely reddish purple coloured solution. The solution 
becomes saturated with cobalt when about 4 c.c. of a saturated 
cobalt chloride solution have been added to 50 c.c. of a solution of 
potassium carbonate containing 100 grams in 100 grams of water. 
This solution on standing very soon begins to deposit crystals of 
the pink double cobalt potassium carbonate K 2 C0 3 CoC0 3 4H 2 0 
prepared by Deville, after some hours the solution has lost most 
of its colour and eventually it becomes almost colourless. The 
double salt behaves very much in the same way as the corre- 
sponding salt of copper, but the solutions are more stable to 
heat : the original solution does not give any precipitate on 
heating. The solutions however undergo a change of colour on 
heating, becoming almost blue, but recover their original pink 
colour on cooling. 
On electrolysing the freshly made purple solution, the coloured 
ion, as in the case of copper, moves towards its anode. Hence 
there is probably a complex ion Co(C0 3 ) 2 which is pink and nega- 
tively charged. This is of interest since in order to account for 
the blue colour of cobalt salt solutions under certain conditions 
it has been assumed that there is a blue complex ion CoCl 4 . 
