112 
MR. B. L). STEELE OX THE MEASUREMENT OF IONIC VELOCITIES 
1 • 
the ratio — tt T ‘v ’ 
U L -4” V 
as Avill be seen, it also measures, in the case of salts of the 
type of sodium chloride, the actual velocity U = xa for any concentration. 
The essential feature of the method consists in the imprisonment of the liquid to 
Ije measui'ed between two partitions of jelly'*" containing the indication ions in 
solutions, thus preventing displacement of tlie liquid during the course of an 
ex})eriment. 
A large number of ditferent forms of apparatus 
have been tiled, and it has been found that for tlie 
measurements of the simplest type of salts, the appa¬ 
ratus shown in hg. 1 is most convenient, t 
In the figure, A and B are two carefully selected 
glass tubes having an even bore, and being both of 
the same area of cross-section; tliey are joined by a 
short piece of tubing of larger diameter. The tube 
B is also provided with a tube E, by means of which 
the liquid during measurement is exposed to the atmo¬ 
spheric pressure. 
Each tube is fitted at either end with the vessel 
C, C and D, D which are carefully ground in. For an 
experiment, two of the vessels (C, C if the indicator 
solutions are lighter, and D, 1) if heavier than the 
measured solution) are taken and filled to a dejitli of 
2 ’5 centims. with a gelatine solution of the indicator 
to lie employed. 
The open ends of the tiilie having been first closed with laibbei' stoppers F, the 
ap)})aratus is filled with the solution and the cells C 0 placed in position, care being 
taken not to enclose any air ])ubbles. Tlie apparatus is then completely immersed 
in a \vater-bath, which is jjrovided with parallel walls of good plate-glass, in (.)i-der 
that the observations inay not be affected Ijy uneven refraction at the surface. 
The electrodes are then placed in position, and the current is started by pouring 
into the cells solutions of the same indicator as that cojitained in the jelly. 
Since tlie boundaries, at first coincident with those between gelatine and aqueous 
solutions, advance shortly after the current is started into the tubes A and B, the 
presence of the jelly can have no influence on theii’ velocities, which are conditioned 
Fig. 1. 
* Attemjits have Ijeeii made to u.se a porcelain membrane instead of gelatine, but it was alwa 3 ’s found 
lliat the liquid was forced through the memljrane in the direction of the anode, proliably on account of 
electric endosmose. 
t A modilication of this form of apparatus may Ijc made b}' connecting the tube with the electrode 
cells by means of rubber tubing, as shown in hg. o. 
