1904-5.] Mr Cameron on the Constitution of Complex Salts. 733 
Apparently the metallic radical in these salts may he any 
univalent or bivalent metal, although in most cases the salts of 
potassium, sodium, and ammonium, or of but one of these, have 
been obtained. With tri valent metals simple salts such as 
Cr 2 (S0 4 ) 3 , and a series such as AlFe(S0 4 ) 3 , are obtained. The 
existence of such compounds as KSrCr(C 2 0 4 ) 3 would appear to 
show that the correct constitutions of these salts are 
(V 
,so 
S0 4 >,r 
'SO. 
and 
S° 
A10 O ^Fe . 
The sesquioxide likewise should be written 
Cr^(ACr, 
X Cr 
and not, as usually, 0 : Cr.O.Cr : 0 . From such considerations 
an alternative formula may be given for the anhydrous salts of 
type B. 2. Thus 
; C 2 0 4 
:0-H H|C 2 0 4 ^ 
g/c"6;Ih H-Of^Cr -> 
X C 2 0 4 -K k-c/v 
It is probable that in all cases where there is water of crystallisa- 
tion the hydroxyl formula is correct, and that treatment of the 
anhydrous salt with water results in the formation of the hydroxyl 
salt before solution is effected, so that the true constitution of such 
anhydrous salts cannot be readily ascertained. 
C 2 0 4 -KK-C 2 0 4 
Further application of the system of types already enumerated. 
There appears to be no reason to doubt that all dibasic acids can 
give rise to complex salts of the kind which have been discussed. 
In connection with the extension of this theory to monobasic 
acids, it is interesting to note the results of Jones and his 
collaborators,* in which they show conclusively by means of the 
conductivities of the respective solutions (as compared with those 
of the solutions of the separate constituents) that in the case of 
the alums, the “ double sulphates ” of the series M 2 'M"(S0 4 ) 2 , 
6H 2 0 , the “ double chlorides, bromides, iodides, and cyanides,” 
* Amer. Chem. J ., 19, p. 83, 22, pp. 5 and 110, 23, p. 89, and 25, p. 349. 
