FLUIDITY AND DIFFUSION

191

always diminishes this region of small velocity, Wilson very
properly attributes this effect to impurity.

Wilson experimented with salol, benzoic anhydride, benzo-
phenone, and azobenzene confined in long glass tubes of varying
diameter. A thermocouple was used to get the temperature of
the solidifying surface, which was of course different in tubes
of various diameters. How well the observed and calculated
velocities of solidification agree can be seen in the following
table for salol.

TABLE XLIX-—THE VELOCITY OF CBYSTALLIZATION OF UNDERCOOLED
FROM WILSON, MELTING-POINT, 42.0, <0= 37.0, a — 0.065,6

Temperature, degrees
t - *0
9
v, calculated
y, observed

35
2
8.77
1.15
1.25

33
4
8.19
2.14
2.5

31
6
7.31
2.90
3.2

29
8
6.49
3.40
3.7

27
10
5.84
3.82
3.9

25
12
5.16
4.05
4.0

21
16
3.90
4.08
4.1

19
18
3.51
4.13
4.1

15
22
2.77
4.00
4.1

Since for all of these liquids the fluidity is as a first approxi-
mation a linear function of the temperature, for salol <p =
(t — 9.8) 2.9, Eq. (66) may be written
v = ap - ftt + 7
where ft = a(9.8 + <o) and 7 = 9-8 at. Thus, whereas it is
possible to express the velocity of solidification as a function of
the temperature only, it is much simpler to express it as a func-
tion of the fluidity as was done in Eq. (66).
Migration Velocity, Conductivity and Transference Numbers.
That the movement of the ions under the action of electrical
attraction should be dependent upon the fluidity of the solution
seems a natural inference from what precedes, and a large num-
ber of researches have been devoted to the elucidation of the
exact relationship. Since the measurement of electrolytic dis-
sociation depends upon this relationship, there can be no question