PROCEEDINGS OF SECTION B. 207 
as of that of the K ions, these keeping pace with one another ; 
and so, at the other end, we have a simultaneous measure of the 
actual equal speeds of Cl and Cr O, But it is important to 
observe that no information is thus gained concerning the 
specific velocities of the coloured ions, or the speeds with which 
they would move if given equal chances with each other and 
with the K and Clions. For, after the first starting of the cur- 
rent, the fall of potential through the tube is not uniform. 
The coloured parts of the jelly, which offer higher resistances 
than the K Cl portion, receive larger shares per unit length of 
the total E.M.F.; and the two coloured portions themselves are 
far from being equal, the blue part offering a much higher 
resistance than the yellow. Thus the observed speeds of the 
four ions concerned in the case are not at all comparable with one 
another, except those of the K and Cl, as already explained ; 
and to make further use of the observations it would be neces- 
sary to have exact measurements of the actual potential slopes 
in the different parts of the tube, and also to take into account 
the unequal ionisation coefficients. There is, however, one way 
of avoiding such complications and still getting some real in- 
formation about the speeds of the coloured ions. This will be 
dealt with in Part III. of this paper. 
In the meantime, one or two other features of the process 
must be briefly mentioned, which are theoretically necessary 
and already proved by experiment (loc. cit.). While the K Cl 
concentration—to retain this salt as a typical example—is pre- 
determined by the operator and retains its value unaltered 
throughout the colourless jelly, however much its length be cur- 
tailed by the progress of the experiment, the concentrations of 
the coloured salts in the other parts of the tube are independent 
of the strengths of the solutions in the flasks and are always 
of smaller value than that of the K Cl itself. Related to this 
is the fact that each ion of the K Cl increases its speed as it 
crosses the boundary travelling towards it, thus spacing itself 
out to suit the smaller concentration of its new partner, and 
also obeying the law of the proportionality of ionic speed and 
potential slope. Further, the total resistance of the tube neces- 
sarily increases as the coloured boundaries advance; and there- 
fore, as the difference of potential between the electrodes re- 
mains practically constant, the current gradually falls off. This 
involves a proportional diminution of the actual velocities of 
all the ions, though relatively to one another they remain con- 
stant. It has been proved experimentally that the theoretical 
numerical relation holds good between the observed intensity 
of the current at any moment and the speed and number of the 
ms which are conveying the electricity across a section of the 
tube. 
Ii a series of experiments be made with different potassium 
