Dissociation and Electrical Conductivity. 439 
pairs of quadrants were first connected together to C and to 
the insulated pole of the battery H. A, D, and J were per- 
manently earthed. The quadrants were then disconnected, 
one pair was kept on the battery, the other pair to C. A 
leak of electricity from © is then shown by a constantly 
increasing deflexion of the electrometer-needle. H is a 
thermometer reading to 360°. 
The insulation was tested when the apparatus was cold, 
and also when A was heated to 360°. The heating was done 
electrically by means of a coil of wire wound non-inductively 
on A and carefully insulated therefrom by a layer of asbestos 
paper. Im series with the heating-coil was an adjustable 
rheostat which enabled the temperature to be varied within 
the limits desired, or kept constant within 1° for a considerable 
time. The substance to be experimented with was placed in 
a shallow metallic dish B. 
Influence of the Nature of the Plate when Salt and Iodine 
are used. 
When the plate B was zinc, Beattie’s results were first 
verified. It was found, however, that the nature of the 
plate, and, in some cases, of the tube A had an influence. 
Up to a temperature of 360°, with salt and iodine placed on 
plate B, the following results with different metals were 
obtained. The electroscope was used in these experiments ; 
the leaves were observed through a microscope having a 
scale in the eyepiece, and the time taken for the leaves to 
fall through the same five divisions was recorded. Cylinder A 
was made of brass. 
| | | | 
_Charge on insulated! Time to leak through 
| Metal. | aaa 5 divisions. ‘ | Temperature. 
| | min. sec. | x 
Aime ;..| Negative ............ 4 0 20 
4 5 112 
oh ap 175 
| 2 50 225 
| be 45 242 
/ 30 255 
| 30 259 | 
| | 25 Peg Pd 
| 
| 
With a positive charge a slightly increased leak was 
obtained on heating; this was irregular. 
Todine vapour was visible at about 100°; its appearance 
did not coincide with an increased leak, sometimes, in fact, 
