the Photoelectric Effect. 591 



currents from platinum at different temperatures, which 

 have been made by F. Deininger*, H. A. Wilson f, and 0. W. 

 Richardson {. These are probably the most reliable deter- 

 minations which have been made, and although they are not 

 in perfect agreement there is no reason why any one of them 

 should be rejected. The values of w in equivalent volts 

 calculated in this way are : — Deininger, 5*03; Wilson, 5*39 ; 

 Richardson, 5" 61. The mean is 5*34 volts or 8*32 x 10~ 12 erg 

 using <? = 4'67 x 10" 10 e.s.tj. The value of v for platinum in 

 the table on p. oS6 is 1*03 x 10 15 sec." 1 whence 



/, = w /v = 8-07xl0- 27 ergsec. . . . (20) 



It will be observed that this estimate of h depends only on 

 thermionic, thermoelectric, and photoelectric measurements, 

 and is quite independent of the estimate & OT = 5*85 x 10~ 27 erg 

 sec. Nevertheless it is very close to the radiation value 

 A = 6'55 x 10" 27 erg sec. The value (20) is, in fact, just 

 about as much in excess of the radiation value as the other is 

 below it. 



The discrepancy between the value of h from iv /v and 

 from the radiation formula might be due either to w being 

 too large or v being too small. It is possible that the values 

 of v estimated from the linear relation between T and v are 

 too small for some unknown reason ; but it is not possible 

 that they are sufficiently in error to account for the whole 

 of the discrepancy. This is shown by the fact that in the 

 case of platinum the photoelectric emission was definitely 

 measurable at X—21 as compared with the value A, = 29*l 

 deduced from the measurement of T r . There is therefore 

 only a possibility of an extension of \ of about 2 units in 

 this direction, and this would only account at most for about 

 half the observed difference. As a matter of fact the tabu- 

 lated values of X on p. 586, although rather higher than the 

 wave-lengths at which photoelectric emission was observed 

 to start, were never very far from them. Although we have 

 made experiments in this direction with most of the sub- 

 stances investigated, we have not been able to satisfy ourselves 

 that there is any certain difference between the tabulated 

 values of X an d the greatest wave-length at which photo- 

 electric emission commences. The tests made, however, are 

 not as delicate or reliable as might be desired, on account of 

 the photoelectric insensitiveness of the materials in this 



* Ann. der Physik, vol. xxv. p. 296 (1908). 

 t Phil. Trans. A, vol. ccii. p. 243 (1903). 

 % Phil. Trans. A, vol. ccvii. p. xxiii. (1906). 



