742 Prof. 0. W. Richardson on the Electr 



on 



combined with the contact potential differences *. As a 

 consequence of this they led to absurdly high estimates 

 of the number of internal free electrons present in these 

 substances t- 



At the time it seemed possible that this discrepancy might 

 be due to the temperature variation of the heat of evaporation 

 being exceptionally large in these cases J ; but the above 

 considerations show that such an explanation is insufficient. 

 It is clear that some other cause must be sought for the large 

 effects observed. In the case of carbon, the admixture of 

 more electro-positive substances would provide an adequate 

 explanation, but it is unlikely that a similar .explanation 

 would account for the behaviour of an extremely electro- 

 positive substance like sodium. In this case, we have 

 presumably to deal with an effect arising from chemical 

 action. 



This position is strengthened by the recent experiments 

 of Pring and Parker § on carbon, and of Fredenhagen || on 

 sodium and potassium. 



The former showed that the currents from carbon could, 

 by purifying the material tested, be reduced to a much 

 smaller value than those which I obtained; whilst Freden- 

 hagen has shown that the currents from the alkali metals 

 are reduced to much smaller values when care is taken to 

 get rid of traces of air or gases which might act chemically 

 on these bodies. 



The values of the currents to be expected from sodium are 

 even smaller than those which I originally estimated, owing 

 to the value of 10 which I used in the case of platinum being 

 too small, as H. A. Wilson showed lateral. The value of the 

 currents from platinum, which is given by Wilson, is 



6-55X10 4 



t ?:=6-9xlO 7 x0ixe ° , . . . (22) 



where x is the current in amperes per sq. cm. of this sub- 

 stance at 6° K. If w r e consider an enclosure containing 

 sodium and platinum in thermal equilibrium and in metallic 

 contact at some point, then if v^ is the number of electrons 

 per c.c. at a point outside the sodium, and v p the corresponding 



* Cf. he. cit. p. 533. t Loc. cit. p. 542. 



t Phil. Trans. A. vol. cci. p. 544 (1903). 



§ Phil. Majr. vol. xxiii. p. 192 (1912). 



|| Verh. d. BeuUch. Physik. Ges. 14. Jahrg. p. 384 (1912). 



U Phil. Trans. A. vol. ccii. p. 243 (1903). 



