642 Profs. Cooke and Richardson on Absorption of 



The mean of the 18 values of <p + 2— (6 — # ) in column 



XII. given by measurements with the standard shunted 

 arrangement is 5*01 volts ; the mean of the 19 values with 

 the filament shunted arrangement is 4*99 volts. Thus there 

 is no certain difference between the results given by the 

 standard shunted and filament shunted arrangements. The 



mean of all the 37 experiments gives cj> + 2— (0 — # ) = 5'00 



volts, or (j£) = 4*7 volts. We see no reason at present why 

 this value should not be correct to 0*1 volt. 



The question whether or no cf> varies with 6 is important. 

 To see if our results give any information about this we 

 have plotted in fig. 2 the experimental values of 



<b+2~(e-e ) 



t 



a 



• r-l 

 O 



I 



^> 



+ 



o 



c3 

 O 



CO 



Fiff. 2. 















© 



©0 otf 



© 



©° a 



© 



_* -©-- 



&>_ © 





© 



5 



&%° o 



© 















/•+ 



/•a 



zz 



2-6 



30 



-^Sc&te of lo3, T 



© m</i'c*/es filament Skunlid 

 o ., standard » 



against the logarithm of the mean value of the corresponding 

 thermionic currents (T). Since the thermionic currents are 

 nearly proportional to e~ const - /d the differences of log T will 

 be nearly proportional to the differences of 6 if the range 

 of 6 is not too large. It will be seen that there is a distinct 



R 

 indication of a slight increase in </> + 2— (0 — ) as log T 



increases. Too much attention ought not to be paid to the 



