564 Mr. G. Jaffé on the Conductivity of 
for each pressure. (The results are represented in figs. 1 
and 2, p. 566.) 
TABLE VY. 
Bulb I. Volume 97°3 ¢.cm.; electrostatic capacity 0°65 em. 
1 scale-division = 0°165 volt. 
we Leak in scale- r y=a,pth,, 
Plage eis! divisions a d, =0°0193, 
; JY} per hour. ae ag ices 
20°5 0°58 0:0285 
42 12 00286 
79 2°15 0:0272 
142 38 0:0268 37 
241 59 00245 a | 
395 8-4 0-0212 86 
538 ody. 0:0217 11°4 
638 1371 0:0205 133 
770 14:8 0:0193 
Tasue VI. 
Bulb II. 
Volume 57°3 c.cm.; electrostatic capacity 1:63 em. 
1 scale-division = 0:2 volt. 
Da cee Number of | Greatest | Leak in meh.) 
observations | deviation | scaie-divi- Leak fam as 2 
see at this from the sions Pressure’ a. 
mercury. i ReesUre | Glee 
pressure. mean. per hour. 2 
33 + 10 p.e. 0:88 00268 
74 tf Bi 53 1-71 0.0231 2°16 
248 7 12s 52 00210 51 
325 3 ae. 6:3 O0-OL95 6:3 
439 6 ile 81 00185 81 
544 3 ee 9-9 00182 9°8 
760 14 23 5, 123 0:0162 
Both series show the same features, the leak being not 
strictly proportional to the pressure. The ratio of leak to 
pressure decreases with rising pressure, as already observed 
by C. T. R. Wilson. It will be seen that the values for the 
leak (y) for pressures from 150 to 600 mm. can be repre- 
sented fairly accurately by a straight line which cuts the axis 
of the ordinates a short distance above the origin, and the 
equation of which is y=ap +b. 
This fact seems to indicate that the ionization is (at least 
