130 
MESSRS. J. HOPKINSON AND E. WILSON ON THE 
drawn in terms of times of contact in seconds. Fio;, 7 gives these curves, which have 
'oeen plotted from Table V. They show that, after a given time of contact, the effect 
of residual charge gradually diminishes as the temperature increases, until only the 
conductivity of the jar for infinite times is experienced. For instance, at about a 
temperature of 250° the table shows that the whole effect of residual charge has died 
away after 1/10,000 of a second. The total capacity of the jar at time t will be 
pi 1 
K + — dt — -— t; where K is the instantaneous capacity which has been found 
J 0 r c X 
by resonance to be = ’0005 microfarad for frequency 2 X 10®. 
= 118,000 divisions of the large slide condenser. 
The curves in fig. 7 have been integrated, and their area up to ‘0028 second, w'hen 
reduced to microfarads and added to K, shows that, for time of contact ‘0028 second, 
the total capacity, which is *000588 at temperature 15*4°, is *00087 at tempera¬ 
ture 145°. This total capacity diminishes as the times of contact diminish, until wm 
get to the results which resonance has shown; and then the capacity of this flask is 
sensibly the same for all temperatures when the frequency is of the order 2 X 10® per 
second. 
Ice. 
Ice was next examined, both in regard to its residual charge and its capacity. 
The residual charge is considerable, and increases as the temperature rises. 
Table VI. gives the residual charge of ice at two temperatures: the higher is 
produced by a freezing mixture of ice and salt, and is about — 18° C. ; the lower by 
placing carbonic acid snow round the beaker, the whole being wrapped in thick felt. 
The apparent capacity depends on the frequency, as shown by the results in 
Table VII. At — 18° the cajDacity is twice as great with frequency 10 as with 77*6. 
At the lower temperature the capacity is greater for frequency 9 than for frequency 
77*6, in the ratio 1*39 to unity. 
The specific inductive capacity of ice w^as next determined, with a high frequency, 
by resonance : it was found to be about 3."^^ Decreasing the frequency to about 
10,000 rendered the method by resonance less sensitive, but it is certain that the 
specific inductive capacity is, for this frequency, of the order 3 rather than 50. We 
conclude that the great deviation of ice from Maxwell’s law is due to residual 
charge, which comes out between frequencies 10,000 and 100. 
Our next step was to determine the resistance c, as in the case of glass, by the 
method shown in fig. 5. The platinum plates, fig. 2, were used, and to observe the 
temperature of the ice a platinum wire of resistance 1‘32 ohms at 0° C. was frozen in 
the ice and surrounded the condenser. Table VIII. gives the results. K the 
capacit}'- as given by the resonance exj^eriments with frequency 2 X 10® was *00022 
g 1 1 
microfarad. Adding’ to this — dt - t, we find that at time *0028 the total 
Jo c 
* Tewing finds 2-85 to 3-36 ; Blonelot 2; Perrot 2-04. 
