﻿802 Mr. Gr. Shearer on the 



of the atomic number, while, if the X-ray energy falling on 

 a metal is kept constant, the number of electrons emerging 

 from the metal is again a linear function of the atomic 

 number of the metal. 



(5) Experiments with Substances other than Metals. — 

 Certain experiments were carried out on the electron emission 

 from salts. Only a few salts were tried, and the data are not 

 sufficiently extensive to warrant the deduction of definite 

 conclusions. The chief salts investigated were As 2 3 , KI, 

 and Rbl. These were chosen in the hope that they would 

 throw light on the electron emission from the elements 

 Arsenic, Rubidium, and Iodine, and thus serve to fill up 

 some of the gaps in the electron emission-atomic number 

 curve obtained from the investigation of the metals. The 

 values found for these salts, with tin taken as 100, were 47, 

 128, and 158 respectively. The values for KI and Rbl 

 were both much higher than is to be expected if the effect is 

 an additive one and the values of the electron emission were 

 such as would result from the interpolation from fig. 2. 

 The atomic numbers of K, Rb, and I are 19, 37, and 53, so 

 that, if the effects are additive, it is to be expected that the 

 values for these iodides would be less than 100, whereas it 

 was found that they were considerably in excess of this 

 value. These salts were placed on the faces of the cube in 

 the form of a layer of small crystals, and it is possible that 

 the effective area under bombardment was thus considerably 

 greater than in the case of a sheet of metal. In spite of 

 this possibility, it seems difficult to account for the high 

 values obtained for these salts. It is interesting to note that 

 the elements potassium and rubidium are both very active 

 from a photoelectric and thermionic point of view. 



A few experiments were made on other salts of potassium. 

 It was shown that the electron emission increased with the 

 molecular weight of the salt. Thus, both the carbonate and 

 the sulphate gave very much smaller value than the iodide, 

 while the effect from the sulphate was greater than that 

 from the carbonate. 



(6) Special Electron Emission. — It has long been known 

 that when a characteristic radiation of an element is excited 

 there is an increase in the electron emission. 



Although, in these experiments, no attempt was made to 

 use monochromatic radiations, still it was expected that it 

 would be easy to detect this special electron emission. In 

 order to test for its effect, observations on two metals — 

 especially copper and tin — were made under widely varying 



