254 Dr. J...W. Nicholson on the Xumher of 



approximately fulfilled, and to be a fair representation of the 

 actual state of things, if the mean free path is properly 

 interpreted. 



The fact that difficulties have been suggested in the theory 

 of emission, of light from metals is not relevant to the present 

 purpose, for, as Wilson has foreseen, it is probable that 

 Jeans\s formula for radiation, based on the hypothesis of 

 equal velocities of agitation, would be obtained exactly if the 

 influence of electronic collisions with one another and with 

 atoms were properly accounted for, even with the Maxwellian 

 velocity distribution. That these collisions are negligible, 

 in a certain sense, in the theory of conduction does not 

 necessarily render them so in that of radiation. But it may 

 be shown in fact, that even neglecting these collisions, a very 

 close approximation to Jeans's formula is found in most cases, 

 so that these suggested difficulties are almost non-existent. 

 Bat we may leave this matter now, as it is foreign to our 

 present inquiry." 



Let us now consider Schuster's values oE N in detail. 

 Neglecting the term in the bracket, as is lawful for all metals 

 whose conductivity is greater than that of lead, 



^ = 31^ i -24 ysiz^oC^ < ^ (38) 



and with = 3 10 10 , and \ = 5'89 10~ 5 for sodium ligl.it, this 

 vields 



p 2 = 183V 2 <r o w<;10- 5 ; .... (39) 

 Schuster gives 



p' 2 = 4:0-6 Y 2 a o v K 10- 5 , .... (4.0) 



and has apparently dropped a factor 4, for this is exactly 

 1/4 of our result. _ The values of p given in his table there- 

 fore require multiplication by 2. In the case of certain 

 metals, a calculation from Drude's constants gives different 

 values for vie, the metals concerned being silver, gold, and 

 copper. For silver we find '66 for 1*66, for gold 1*03 for 

 2'()6, and for copper 1*68 for 1*60. 



Schuster's calculation has been repeated, with the adoption, 

 however, in possible cases, of the now accepted measurements 

 of conductivity of Jager and Diesselhorst *, as the most 

 accurate. They are the values oE the conductivities at 18° C. 

 The results of this repeated calculation are shown in Table I., 

 the complete formula (37) being used in necessary cases. 

 For cobalt, antimony, magnesium, and lead the conductivities 

 are taken from Schuster's table. 



* Wissen. Abhand. der Physikal Techn. Reichanstalt, iii. p. 2G9 (1900). 



