280 Proceedings of Royal Society of Edinhurgh. [april 2, 
in tlie percentages of these two metals, calculated from theory for 
the two salts of ratios, 9 : 2 and 2 : 1 is very marked, and the 
results of the analysis agree very well with the theoretical percen- 
tages of the formulae assigned. It is in accordance with the 
behaviour of isomorphous mixtures, that when two salts are in 
solution together that which is less soluble has a tendency to 
crystallise out at first in excess of the other. Thus it is that in the 
case of ISTo. I. we have a salt formed having the ratio of the two 
metals Co : M = 1 : 2. So also in the case of copper-cobalt-potas- 
siuin sulphate, the first two crops have the ratio Cu : Co = 5 : 4. 
The withdrawal of a proportionally large amount of copper-potas- 
sium sulphate having altered the ratio of the metals in the mother- 
liquor, the influence of the greater solubility of the cobalt-potassium 
sulphate was counteracted by its presence in a relatively larger 
quantity; and the next crop of crystals consisted therefore of a salt 
of Volil’s type, in which the ratio of the metals Cu: Co = l : 1. 
Preparations X. and XII. show striking instances of the influence 
of solubility. The ratios in both are nearly 1:12, the greater 
number referring in each case to the less soluble salt. The results 
of the analysis of these samples I do not take for more than an 
indication of how the respective sulphates tend to behave when 
taken in equivalent proportions. To study the behaviour of these 
sulphates further, preparations XI. and XIII. were undertaken. 
In each of these two cases the more soluble salt was taken in four 
times the equivalent proportion of the less soluble one. Under 
these circumstances, salts were obtained having the formulae assigned 
to them in the table. 
The difficulties in the way of discriminating between a homo- 
geneous crystallised salt and a mixture consisting of two or more 
salts in variable proportions are undoubtedly great, Rammelsberg 
says — ‘‘ The analysis of a certain number of crystals is no guarantee 
for the constitution of individual crystals . . , . The result in general 
must be regarded as an expression for the mean of one entire crop. 
.... Unfortunately, I could never analyse any of the crystals singly 
on account of their diminutive size ” (loc. cit.^ page 329). Von 
Hauer attempted to get over this difficulty by analysing individual 
crystals which he was fortunate enough to obtain sufficiently 
large. 
