IN AQUEOUS SOLUTION, AND THE EXISTENCE OF COMPLEX IONS, 
141 
Thus the equation in the foiin given l^y Massox no longer holds good for solutions 
which contain complex ions, a conclusion which is verified by the irregular values 
C , ' . 
for the ratio --— — ^ shown for these salts in Table V. 
Xne (U + \) 
The conclusions arrived at by Kohleausch, and given in the early part of this 
pajDer, as to the movements during electrolysis of a portion of an electrolyte which 
differs from the remainder in concentration, these movements being conditioned by a 
change in transport number, receive a simple explanation by the assumption of 
complex ionization. Let us imagine such a solution having initially a section which 
differs from the remainder in concentration, and first let us assume no complexes to 
be present, that is, p constant. Then since only the simple ions carry the current, 
that portion of the salt which is not ionized remains stationary (neglectiirg diffusion), 
and no movement is brought about. Next let us assume complexes to be present, 
that is, p> variable; then, in addition to the motion of the completely ionized salt, a 
portion at least of the remainder of the salt is carried liy the current, and movements 
of the section take place. If the complexes are cations, that is, p decreases witli 
increasing concentration, the movement is in the positNe direction ; if, on the other 
hand, p increases, or the complexes are anions, in the negative direction, in complete 
ao-reement witli Kohleausch’s conclusions. 
o 
This behaviour of such a section explains at once the change in p that is found 
when the velocities of the two boundaries are measured ; for, assuming that the 
margins are those between indicator and simple ion, there occurs simultaneously and 
independently the migration of the complex, by means of which the whole column of 
salt solution (cations and anions) is carried along and in the positive or negative 
direction, according as the complex ions are cations or anions, and hence the apparent 
velocity of the one boundary is diminished and that of the other increased. The 
influence of the presence of such complexes on tlie velocity of the margin between 
two solutions is a question of considerable importance, and one tliat is worthy of strict 
mathematical investigation. It is, however, not considered by either Kohleausch or 
Webee in the papers previously cited. 
It is possible that, at the anion boundary, where the indicator is following two ions 
which differ in velocity, the concentration of the solution within the margin may 
become slightly altered. 
Such a change undoubtedly occurs at the anion boundary of solutions of magnesium 
sulphate and copper sulphate, and the change is in the direction of a diminution of 
concentration at this point. This corresponds with an increased resistance, and 
accordingly a higher j^otential fall, and hence a greater velocity for this margin, 
a result which would also help to explain the increase in p with more concentrated 
solutions. 
Experimental evidence of the changes referred to has been obtained in the following 
manner :— 
