192 



Dr. B. Hodgson on Heating Effects at 



Fig. 1 shows four results taken from Table I. Three of 

 the curves are moved bodily upwards for the sake of clear- 

 ness. Energy supplied to the anode per second is plotted 



Fia-. 1. 















■/ 











/■Srn-n, 











X to'*' 1A 



V 



























/ J 



'■2/irr, 







ry 





/ 



/ 



'. 









jtMA. 







/ 



/ 











*<j£C<r7 



cL 





f 



/ 











A 







7 



( 











1 







/ 







l/ZZS'^y 



' 





/ 





/ 







y 









/ 



/ 









X6Sr,^. 





/ 



/ 



A 



/ 















// 







■cu. 



T*y-&nt 











■O0~J a^tn^S 



against current for constant pressure. A linear relation 

 exists at the pressures used. For lower pressures the curves 

 are of a different type, and fig. 2 gives some of these, which 

 all possess the same general shape. For small currents 

 proportionality exists between current and energy, and then 

 as the current increases the energy communicated to the 

 anode actually decreases and falls to a minimum value, after 

 which a quick rise is obtained. For currents smaller than 

 0*001 amp. no values were obtained, but the probable 

 courses for the 0*28 mm. and the 0*35 mm. curves are 

 shown by the dotted parts. The curves in fig. 1 would 

 probably show the same form could currents large enouoh be 

 obtained. 



If certain assumptions be made, the average enero-y per 

 ion can be obtained. If n negative carriers with only one 

 charge strike the anode per second and do not ionize there, 

 then ne — i, where i is the current and e the charge per ion. 

 Then the energy given to the anode per second is neV where 

 V is the mean potential through which the ion falls, and 



neY' = iV. 



