FLUIDITY AND DIFFUSION 197

• In the same way the effect of the addition of sucrose to lithium
chloride solutions would be explained by an increase in one or both
of the ionic'volumes due to uniting with the sucrose molecules.
That such a hypothesis is not improbable, it is well to add that
C. H. Gill has found that sucrose does form crystalline com-
pounds with the halides of sodium and ammonium. He did not
find that lithium chloride forms such a compound but there may
be a sufficient tendency to unite in solution to explain the effect
which seems to be peculiar to sucrose. Glycerol, although
highly viscous like sucrose, gives values of m which are unity,
according to the determinations of Massoulier (1900) as calcu-
lated by Green.

In conclusion, it may be added that there is no connection
between the conductivity and the fluidity of a colloidal solution
of gelatine, as demonstrated by Griffiths (1896) (see also
Ltideking (1889)). The reason for this peculiarity lies in the
heterogeneous character of colloidal solutions as will be more
fully discussed later. Schweidler (1895) has also shown that
there is no relation between conductivity and fluidity in mercury
and certain amalgams.

The Transference Number.—The transport number nA is
expressed by the equation

If the equivalent conductances of the different ions change with
the fluidity at different rates, the transport number must be also
a function of the fluidity. We have the two equations

and

A A /M"*

AB = AooB I----I

x^oo/

whence, according to Washburn (1911)

where N&A is the transport number of the anion at infinite dilu-
tion and m is the exponent of Eq. (62) for the salt.