1904-5.] Mr Cameron on the Constitution of Complex Salts. 727 
It has been pointed out from other considerations that the 
tendency in such solutions is for the formation of crystals in 
which the metallic radicals have a simple ratio. 
I have endeavoured to determine the molecular weight of 
Gregory’s salt by means of the freezing-point method, hut this 
compound presents some anomalies in that connection ; these are 
at present under investigation, and may form the subject of a 
future paper. 
Rosenheim and Koppel * have determined the conductivities of 
a considerable number of complex oxalates, and the result 
/ x 1024 — A t 32 = 33 - 3 
for Gregory’s salt would lead to the conclusion, from Ostwald’s 
empirical rule, that this is a salt of a tribasic acid. 
I conclude, therefore, that in the case of the chromoxalates and 
similar salts there is at present no special reason for using doubled 
formulae, and that until such reason is put forward I am justified 
in using the simple empirical formulae in every case. Should more 
weighty proofs he put forward for the doubling of these formulae, 
the reasoning in the following pages is not invalidated, though the 
formulae are rendered more cumbersome. 
In considering the constitution of salts, it must be remembered 
that in crystalline solids there exist molecular complexes whose 
nature is little known, and that solution breaks these up into 
simple molecules, often ionising the latter. Such considerations, 
then, are really concerned with the non-ionised molecules in solu- 
tion, and it is possible that many crystalline solids are outwith 
the subject, as their complexes may be incapable of existing in 
solution. 
In my subsequent remarks I have not taken into account water 
of crystallisation, other than water of constitution, although it may 
he pointed out that certain well-defined regularities exist. These 
are most marked in the case of potassium compounds. To give hut 
one example, salts analogous to the blue potassium chromoxalate 
almost invariably, like it, contain three molecules of water. 
Zeit. fur anorg. Chem., 21, p. 17. 
