86 Mr. F. S. Spiers on Contact Electricity. 



as is evident from the above readings, such was not the case; 

 there seemed to be no tendency for the voltage to settle down 

 to a final steady value. However, a careful examination of 

 the above results and of others not given here, convinced me 

 that there is a law underlying these results in spite of their 

 apparently contradictory nature. It seems that, on the whole, 

 readings that were taken as soon as possible after the heating 

 (the tube, however, being cold or almost cold) were more 

 positive (or less negative) than those taken some hours or on 

 the next day after the heating. In order to test this suppo- 

 sition, when I next repeated the operation of heating, I 

 took time -readings of the voltage, starting as soon as possible 

 after the heating. There is really a double time effect taking 

 place, the one superimposed upon the other : firstly, an 

 ordinary temperature effect ; this ceases when the tube is 

 quite cold or soon after ; secondly, the time effect particu- 

 larly referred to here, which goes on, as will be seen, for 

 several days, long after direct temperature effects have dis- 

 appeared. In consequence of this the time-readings taken, 

 say within the first two or three hours after heating, teach 

 us nothing definite. 



The tube was kept red hot for half an hour from 2.20 to 

 2.50 p.m., and then the asbestos-lined copper shield was 

 immediately removed so as to allow it to cool as quickly as 

 possible. At 3.30, when the apparatus felt almost cold, the 

 P.D. was 



0-18 volt. 



At 3.50 it was 0'04 volt. 



At 4.25 it was -0'12 volt. 



At 5 -0'15 volt. 



At 8 -0-20 volt. 



The next day at 10.30 a.m., -0'29 volt. 



After two more days, at 10.30, —0*60 volt. 



The temperature-E.M.F. curve becomes asymptotic at 

 about -0-60 volt. 



On repeating the operation by again heating and taking 

 time-readings, I obtained very similar results. In this case 

 the P.D. after twenty-three hours was 



-0-28 volt, 



and I then started exhausting the apparatus of hydrogen 

 (although the equilibrium value, — 0'60 volt, was not yet 

 arrived at) with the intention of again letting air into the 



