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



that followed were similar to those observed in the old 

 apparatus (§ 8). Even after heating in this way for rive hours 

 the pressure was about the same as at the beginning of the 

 operation (it varied from ^ millim. to ^ millim.), showing 

 that after prolonged heating condensed air was still given off 

 from the inside surfaces of the apparatus. This did not, of 

 course, necessarily imply that there was still a large quantity 

 of air left sticking to the metal plates, for as the heat slowly 

 travelled up the tube by conduction air would be gradually 

 driven off its upper surfaces, and this would continue until 

 the whole of the apparatus (and even the pump) became 

 strongly heated. The source of heat was very gradually 

 removed, the pump being vigorously worked the whole time, 

 until when the tube was quite cold the pressure was 25V0 

 millim. The P.D. between the plates was then 



0-28 volt. 



(I could never get a satisfactory reading until the tube was 

 quite cold, for on insulating the platinum plate and its con- 

 nected quadrants there was always a considerable leakage of 

 electricity into the plate, possibly by convection from the sur- 

 rounding gaseous molecules.) 



After leaving the apparatus for 15 hours the pressure had 

 risen to ■£$ millim., and then the P.D. was 



0-31 volt. 



On exhausting to goVo millim. (without further heating) 

 the P.D. was still 



0-31 volt. 



Air was then admitted into the tube. The reading remained 

 unaltered even after several hours, but after leaving the tube 

 open to the atmosphere for three days, the P.D. had fallen to 



0'02 volt. 



The aluminium plate was then taken out and very gently 

 rubbed once or twice with fine emery-paper. The P.D. 

 rose to 



056 volt, 



It was again rubbed with emery, after which the P.D. was 



0-72 volt. 



It is thus clear that at least the greater part of the above 

 diminution of the contact effect was due to some change of 

 surface of the aluminium plate, most probably oxidation, 

 since it was a permanent effect and only removed by cleaning 

 the plate, and also because almost the same phenomenon took 



