﻿Electrical Changes induced by Ultraviolet Light. 417 



used for several experiments of short duration, save a 

 " tiring " Curve V., which lies between II. and III. Similar 

 " tiring curves " were obtained for zinc, using the reflexion 

 method (fig. 2), in which no charge was put on the plate. 

 To explain the foregoing experimental results we mast recall 

 the idea, postulated in the introduction, that a metal is com- 

 posed of a large number of electrons, of which a few, pre- 

 viously termed by us "metallic corpuscles/' and corresponding 

 to its valency, differ from the residue. The explanation will 

 be clearer if we consider the experiments as made by the 

 reflexion method, in which the plate was not charged. When 

 ultraviolet light falls upon a plate of metal, e* g. Mg, it effects, 

 as before stated, an expulsion of electrons; this splitting off 

 proceeds at a continually decreasing rate, until a certain 

 point indicated in the curve is reached : at this point a con- 

 stant rate of expulsion of electrons is obtained which must 

 be attributed to the loss of the first of the " metallic cor- 

 puscles," during which we have a constant electric pressure, 

 just as we have the constant vapour pressure of various 

 hydrates in the case of dehydration of certain crystallized 

 salts, e. g. CuS0 4 5H20. The steady rate of discharge con- 

 tinues until the metallic electron has been expelled from the 

 whole of the atoms at or near the surface of the plate. Then 

 we get a further splitting off of the other electrons until we 

 come to the second break in the curve where the second 

 metallic electron is expelled, the constant rate of discharge 

 being explained as before. We obtain repetitions of these 

 breaks so long as there are metallic corpuscles still combined 

 in the atoms ; when the last of these has been expelled we 

 have a curve which exhibits a continuously decreasing rate 

 of discharge, this means that the metal is undergoing a further 

 disintegration at a slowly decreasing rate. Aluminium, with 

 its five or six breaks, does not seem to fall into line with the 

 above hypothesis ; the additional breaks, however, may be 

 due to the formation of definite electron complexes, less, or 

 even more, complicated than the complex we know as alumi- 

 nium. If we now consider the successive curves obtained 

 for magnesium, we have here the breaks approximately in 

 the same places, but they are of much shorter duration, and 

 there is a very much increased rate of tiring in each succes- 

 sive curve. The increased rate of tiring can only be explained 

 by the fact that the magnesium has become permanently 

 changed, at least on the surface. The persistence and de- 

 creasing length of the delays in the successive curves show 

 that, although the metallic corpuscles are removed from the 

 surface layer, yet there is a slow diffusion of electrons from 

 Phil. Mag. S. 6. Vol. 12. No. 70. Oct. 1906. 2 E ' 



