Dissociation and Electrical Conductivity. AAT 
is more easy to produce than it is when the ionization is 
confined to a single layer near one electrode. 
It can be shown, on certain assumptions*, that the current 
should vary as the square of the potential-difference between 
the electrodes when the ions are produced in a narrow space 
near an electrode. When the current is carried by positive 
ions, it was found in our experiments that it is expressed with 
considerable accuracy by a formula of the type l=av+ bv’, 
where I[=current, v=voltage, and a and 6 are constants. 
The points on the diagram marked with a cross were calcu- 
lated from such a one, The other curve of fig. 6 is nearly 
parabolic. 
If the conduction caused by heating a mixture of zinc, 
salt, and iodine be due to the formation of ZnCl, one way of 
identifying this would be to measure the rate of decay of the 
current produced by ZnCl, when kept for some time at a 
constant high temperature, and show that the law is the same 
as for the mixture. 
The conduction obtained with ZnCl, alone was very ir- 
regular in amount; it had decreased by less than 10 per cent. 
when kept for 34 hours at 360°. After standing for some 
hours in the apparatus, the current was largely increased on 
reheating, due probably to the absorption of moisture. The 
experiments are being repeated and extended with substances 
that have been carefully dried. 
Experiments with a Miature of Zine, Salt, and Iodine. 
# About + gram of a mixture of zinc, salt, and iodine was 
placed on the earthed plate of the condenser and heated, the 
current due to 200 volts being noted at different temperatures. 
One set of readings is given below. 
C charged to —209 volts. 
| Temp. | Leak per min. || Temp. | Leak. 
5 Semaine a 2s. =o | en ae ee Pe 
LS oe | 3 Wie ae il. 28 
95 | Irregular. Todine-|| 342 | 52 
| vapour coming off. || 346 80 
| 240°-"| 3 3515 | 120 
| 302 3 304 | 155 
312 2 307 210 | 
323 4 359 250 | 
som 13 | | 
* “Conduction through Gases,’ p. 175. 
