390 Photo-Electric Effect of some Compounds. 



much smaller than with the mercury salts. On the inter- 

 pretation suggested above of an increasing photo-electric 

 effect for salts, it appears that lead iodide is decomposed 

 very slightly by light. It is possible that a photo-electric 

 test of this kind may be a much more sensitive test for 

 decomposition by light than any of the tests hitherto used. 



Putting aside the copper oxide, the magnitude of the 

 photo-electric effect and the rapidity with which it sets in 

 run more or less parallel with the heat of combination of: the 

 salt. Thus the most stable of the salts ZnCl 2 (when dry) 

 shows no photo-electric effect at all, and the chlorides show, 

 in general, a smaller effect than the iodides. 



Several experiments show clearly that, if the distilled salt 

 is not exposed to ultra-violet light for some hours, then only 

 the small initial leak is obtained on the first exposure, proving 

 that the increase in the effect depends upon exposure to 

 light. 



Summary. 



1. An investigation has been made of the photo-electric 

 effect in ZnCl 2 , P 2 5 , Pbl 3 , Hgl, Hgl 2 , HgCl, HgOl 2 , FeCl 3 , 

 Sbl 3 , BiCI 3 , the oxides of Ba and Cu, and anthracene. 



2- Dry Zn(Jl 2 and P 2 5 show no photo-electric effect. 

 There is practically no effect for FeCl 3 , and only a very 

 small one for Pbl 3 after exposure to ultra-violet light. The 

 other halogen salts show well-marked effects after exposure 

 to ultra-violet light. 



3. The initial effect is either zero or very small for the 

 halogen salts but, after exposure, increases in many cases to 

 a large value. (Sbl 3 is an exception to this rule.) 



4. The explanation suggested is that the salts themselves 

 are not photo-electric with wave-lengths longer than X1849, 

 but that the light first of all decomposes the surface, and then 

 acts on the metallic element in the ordinary way. The more 

 stable the salt, the less is this effect. 



5. The photo-electric effect in anthracene is produced 

 entirely by wave-lengths shorter than X2002. The maximum 

 emission velocity corresponding to X1849 is '87 volt, 



6. ZnCl 2 and P^0 5 have no photo-electric effect when dry, 

 but after contact with moist air they show a distinct effect. 

 This effect cannot be accounted for by the photo-electric 

 activity of water which, if it exists, is much too small. 



I have pleasure in thanking Professor Sir J. J. Thomson 

 for his interest and encouragement during this investigation. 



The spectrograph used in some of these experiments was 

 obtained by means of a Government Grant, through the 

 Royal Society. 



