Voh. 6, 1920 
CHEMISTRY: W. D. HARKINS 
603 
It is of interest in this connection to note that while in the papers 
previously cited the hypothesis is presented that the ionization is com- 
plete, and while according to the commonly accepted theory the ioniza- 
tion of such electrolytes, when uni-univalent, is about 0.85% at tenth- 
normal concentration, G. N. Lewis, and Lewis and Linhart,^ have ad- 
vanced the idea that the corresponding value is 68.4% (as calculated 
from the data of Hall and Harkins on sodium and potassium iodates). 
This they call the thermodynamic or corrected dissociation. It may, 
therefore be considered that such salts are 100% polarized or ionized, 
85% electrolytically dissociated, and 68% thermodynamically dissociated. 
All of these conceptions are probably of use, and no one of them refers 
to the same feature of the phenomenon as any of the two others. 
The confusion arises partly, as has been stated, from the attempt to 
consider a kinetic or a statistical problem as giving an average or mean 
which has no meaning from the standpoint of the condition in the solu- 
tion, but has a considerable value as the basis for calculations by the use 
of empirical methods, as for example calculations of solubility or of electro- 
motive force. Those who rally to the support of the complete ionization 
theory are doing so largely to do away with these mean values, and their 
endeavor is to introduce a statistical theory which shall be more closely 
descriptive of the phenomena. However, I wish to point out that to go 
over completely to this point of view at the present time, before the 
statistical theory has been worked out in an entirely satisfactory manner, 
is to lose some of the useful features of the older theory. 
Only one of these features will be discussed. In 1911 I showed^ that 
the effects of salts upon the solubility of unibivalent salts, are such as to 
almost inevitably lead to the conclusion that unibivalent salts ionize, 
or dissociate in steps, to form what were then called intermediate ions. 
According to Milner's general nomenclature these might be called asso- 
ciated ions. A calculation made by G. N. Lewis showed^ that the trans- 
ference and solubility results on this type of salts, could only be brought 
in harmony, on the assumption that the intermediate ion is so largely 
separated into its component ions, and that while it moves together under 
the influence of the electric field, it offers practically the same resistance 
to motion through the solution as if it were completely separated. Never- 
theless, that there is such a union into an "intermediate" or associated, or 
possibly "complex" ion is clearly indicated by the phenomena. Noyes 
and Maclnnes, though they do not consider this problem, seem to take 
the point of view that such relations can be accounted for by changes in 
the activities of the ions. However, the solubility relations of such salts 
could have been predicted from the standpoint of the intermediate-ion 
theory without any further knowledge of the relations other than those 
of the ordinary theory of solutions; but the activity concept, like other 
thermodynamic relations, is not specially fruitful in this particular sense. 
