Photo-Electric Effect of some Compounds. 389 



imperfect insulator, for, after an exposure of several minutes, 

 the leak was reduced to about half of its original value. On 

 increasing the accelerating potential by a few volts, the leak 

 increased to approximately its original value, showing that 

 the diminution was not due to fatigue, but to the surface 

 acquiring a positive charge, and thus reducing the effective 

 accelerating potential almost to zero. It may be mentioned 

 here that similar tests for poor conductivity in the other 

 compounds were frequently tried. Their conductivity, how- 

 ever, appeared to be sufficient for these experiments. 



Discussion of Results. 



It was noticed, at the end of each experiment on the 

 mercury salts, that the surface had a more or less grey 

 metallic lustre. The appearance of the exposed part of the 

 surface was quite different from that of the unexposed part. 

 It is known that mercury salts are decomposed by ultra r- 

 violet light. In the case of the chlorides, there is no 

 appreciable emission of electrons until after an exposure of 

 thirty seconds or more. Then the effect increases with ex^- 

 posure, apparently showing that the effect of the light is to 

 decompose the salt first of all, and then to act on the mercury 

 surface as upon ordinary metals, This effect is more marked 

 in the case of the iodides than in the case of the chlorides. 

 This is readily explained by the fact that the iodides are 

 more easily decomposed than the chlorides. Jt is difficult 

 to say whether the small effect obtained in the f;rst 

 20 seconds with the iodides is a real effect with the unde-? 

 composed salt, Analogy with the chlorides and other salts 

 suggests that the decomposition sets in too rapidly for the 

 initial stage, during wjaich there is no emission, to be detected. 

 The photo-electric behaviour of the antimony iodicle appears 

 to be an exception to the general rule. 



In a paper recently published, Heboul * finds that the 

 photo-electric leak from the halogen salts of copper and silver- 

 increases for a considerable time with exposure to light, 



It is worthy of note that, in many cases, the leak decreases 

 slightly after the salt has been kept in the dark, indicating 

 perhaps that a partial recombination is taking place. Com-? 

 plete recombination is impossible as the charcoal absorbs 

 any chlorine or iodine vapour which gets away from the 

 surface. 



It is not known whether lead iodide js decomposed by 

 light. There was no trace of discoloration of the surface 

 even after prolonged exposure. The leak, however, was 



* Eeboul, Le Radium, x. p. 155 (1912). 

 Phil Mag. S. 6. Vol. 24. No. 141. Sept, 1912, 2 D 



