IN AQUEOUS SOLUTION, AND THE EXISTENCE OF COMPLEX IONS. 
113 
only by the nature of the preceding ion and by the potential fall, provided certain 
conditions, which will be described immediately, are fidfilled. 
It has been already pointed out that it is not necessary that either the measured 
or indicator solutions should be coloured, a perfect boundary being rendered visible 
in most cases by the difference in refractive index of the two solutions. 
The occurrence of such a boundary has been previously noted by Lenz (‘ Mem. 
Acad. St. Petersburg,’ vii., vol. 30, No. 9, 1882) in the case of cadmium chloride 
following sodium chloride, and Bein (‘Zeitschrift Phys. Chem.,’ vol. 27, 9, 1898) also 
refers to the same kind of boundary for the same pair of solutions. 
The Production and Maintenance of a Good Boniidari/. 
In what follows, by a good boundary will be understood, one which moves with 
constant velocity under a constant potential fall, or whose change in velocity is 
proportional to the change in potential fall.t 
For the production and maintenance of such a boundary, the following conditions 
are necessary :— 
1 . The indicator ions must have a specific velocity slower than that of the ion 
to be measured. 
2 . The indicator ion must not be such as to react chemically on the solutions to 
be examined. 
3. During electrolysis the cell solutions must not give rise to any species of ions 
which would move faster than and overtake the measured ion, thus altering the 
potential slope within one or both boundaries. 
4. The specifically lighter solution should lie over the heavier. 
5. The indicator should have a resistance not very much greater than that of the 
solution it follows. 
6 . The potential fall should lie within certain limits, which depend on the nature 
of the solutions forming the boundary. 
In addition a tube should be selected ol' such a size that with the required 
potential fall the total current does not exceed 0'03 ampere. If this limit is exceeded 
there is considerable danger of the jellies being melted. It has been found, 
however, that in a tube of smaller sectional area than 0'08 sq. centim. there is great 
difficulty in detecting the position of the margin. 
Of these conditions, the first three have already been given by Masson for the 
measurement in gelatine. In aqueous solutions a refraction boundary may occa- 
* This refraction margin may be shown to a large audience, or as a lecture experiment, l)y means of 
projection lantern, when it is seen on the screen very clearly and distinctly. 
t The best test of a pair of boundaiies lies in the fact that if their velocities U and V l»ear a constan 
ratio to one another they are probably both good, if the ratio is not constant one or both are bad, and 
is easy by plotting the velocity curve to see whether the boimdary fulfils the l equired conditions. 
VOL. CXCVIII. — A. Q 
