Mr Wood and Mr Jones, On Complex Carbonates. 175 
function of the bicarbonate in the solutions of Soldaini and Ost 
is to ensure that the copper and potassium in solution are com- 
bined with or in equilibrium with C0 3 ions and not with hydroxyl 
ions. 
The solutions are of a very deep blue colour, slightly darker 
than equivalent alkaline solutions of copper tartrate and not quite 
so dark as equivalent solutions of copper in ammonia. The 
copper is precipitated from the solution by ammonium sulphide 
or potassium ferrocyanide but not by caustic potash in the cold ; 
potassium iodide has no effect on the solution but potassium 
cyanide decolourizes it. 
On electrolysis copper is deposited at both poles, this 
suggested the existence of complex ions in solution. To test 
this, experiments were made on the movement of the coloured 
boundary between a solution of potassium carbonate and a 
solution of the double copper carbonate in potassium carbonate. 
The blue boundary moved towards the anode on passing a current 
through the solution : the kathode junction remained sharp while 
the anode junction became diffuse. When a solution of potassium 
chloride was used instead of the potassium carbonate the same 
result was obtained, except that a slight precipitate of basic 
copper carbonate was formed at each junction. 
It is therefore probable that the salt is to be represented as 
K \ 
CJu/ 
r/ 
C0 3 
C0 3 
as suggested by Reynolds and that it dissociates into the ions 2K 
and Cu(C0 3 ) 2 . The anion Cu(C0 3 ) 2 decomposes or dissociates 
partially into Cu and C0 3 ions, since the solution contains enough 
Cu ions to reach the solubility product of copper sulphide or 
copper ferrocyanide on the addition of ammonium sulphide or 
potassium ferrocyanide respectively. 
Some experiments were also made in which the solution was 
saturated with potassium carbonate, bicarbonate and basic copper 
carbonate or double salt. The phenomena were more complicated 
and were not fully investigated, but the solutions always con- 
tained less copper than when no potassium bicarbonate was used. 
In no- case did we observe the formation of the greenish blue 
tables of K 2 C0 3 CuC0 3 4H„0 described by Reynolds, but we ob- 
served the formation of the silky blue needles of K 2 C0 3 CuC0 3 H 2 0 
during some experiments in which Groger’s salt was used as a 
source of copper instead of the basic carbonate or the double salt. 
Groger’s salt, as stated previously, is described as 
K 2 C0 3 .8Cu 0.7C0 2 .17H 2 0 or 4KHCO s . 8CuO . 5C0 2 . 15H a O, 
