382 Dr. A. LI. Hughes on the 



the emission of electrons from the surface of the disk can be 

 investigated. When it is desired to cover the disk with a layer 

 of distilled salt, the disk is lowered until it is near the furnace F. 

 The furnace consists of two quartz bowls, one of which fits into 

 the other. Between the two bowls is a broad strip of Pt or 

 Ni foil which can be heated to redness by an electric current. 

 The substance to be distilled is placed inside the smaller 

 quartz bowl. The distillation is effected in a liquid air 

 vacuum, and the disk, when covered with a suitable layer of 

 the distilled substance, is raised into position where it is 

 exposed to the light. The pressure of the residual gases never 

 exceeds "001 mm. during the process of distillation. 



The source of ultra-violet light was a mercury arc in 

 quartz glass, the shortest wave-length emitted by this lamp 

 being X1849*. The lamp was always placed at a distance of 

 30 cm. from the apparatus. 



Experimental Results. 



These results were usually obtained with an accelerating 

 potential of + 2 volts on the case. Certain tests, described 

 later, were nsed when it was suspected that the conductivity 

 of the substance was so small that its surface became charged, 

 and therefore at a different potential from that of the disk. 

 Except in the case of BaO, the conductivity of the salts was 

 good enough to enable the experiments to be carried out 

 without any trouble of this kind. The leak on the first 

 exposure to the light was measured in as short a time as 

 possible, since the magnitude of the effect grows with the 

 time of exposure. An appreciable time, however, is required 

 to determine the rate at which the gold leaf of the electroscope 

 moves, and hence it is that the time during which the leak 

 was measured could not be reduced below about 20 or 

 30 seconds. In the following table, for example, the leak 

 on the first exposure was obtained by observing the rate of 

 motion of the gold leaf during the first 40 seconds. (In the 

 second table, the leak during the first minute's exposure was 

 measured in three parts and the growth of the effect is 

 clearly shown.) The leak was again measured over a period 

 of 30 seconds at the end of 5 minutes' exposure to the light. 

 The surface was then shielded from the light for 5 minutes, 

 and the leak was once more measured when the light was 

 directed on to the surface at the end of that period. The 

 unit of photo-electric leak throughout the paper is the same, 



* Hughes, Proc. Camb. Phil. So?, xvi. p. 428 (1912). 



