PROI)IHT.1> |:v HOT I'LATIM'M IN DIFFERENT OASI x 



25 



-9 



-10 



-11 



"6 _ 



13 



-14 



-16 



-16 



\ 



600 TOO 800 900 1000 1100 



Scale of 10* + 6 



Fig. 10. 



Expressed in this way, the numbers can at once be compared with the heats of various 

 purely chemical reactions. The best series of experiments in oxygen (at 1'5 millims. 

 pressure) gave for the work required to set free a corpuscle o_ = 13'55 x 10 4 calories. 

 This is in good agreement with WILSON'S value 13'11 x 10* for the negative leak from 

 hot platinum in air at a low pressure.* The value of the work required to set free a 

 positive ion was found to be tu + = 3'04 x 10* calories. The heats of the most intense 

 chemical actions involving only one valency range around 5 x 10 4 calories, so that the 

 energy required to liberate a positive ion is somewhat smaller, and that required to 

 set free a negative ion considerably bigger, than the greatest amount of energy set 

 free in any known chemical reaction. 



7. Uncontrollable Variations. 



We have seen (p. 16) that the positive leak in oxygen, particularly at high 

 pressures, continually varies in an erratic manner, even when all the controllable 



' Phil. Trans.,' A, 362, vol. 202, p. 269. 



vol.. i -CVII. A. 



