252 Mr M c Glelland, On the Conductivity of Gases 



air or oxygen. We saw above that there were reasons for suppos- 

 ing that in the arc itself the very great initial difference in the 

 masses of the positive and negative carriers was longer maintained 

 than in air or oxygen, and this result obtained from actual 

 measurement of the velocities would point in the same direction. 



(18) At this stage in these experiments the arc was discarded 

 and an incandescent wire used to produce ionisation and render 

 the gas conducting. 



There are many points of resemblance between the conduc- 

 tivity of the gas and its electrification as produced by the arc and 

 by an incandescent wire, and hence the results obtained with the 

 wire may be given in continuation with those recorded above for 

 the arc. 



To adapt the apparatus described at the beginning of this 

 paper to this work it is only necessary to substitute a piece of fine 

 platinum wire a few centimetres long for the arc, and raise it to 

 the required temperature by passing a current through it. 



We shall first consider the case when the wire is heated in air. 



Incandescent wire in air. 



(19) By gradually increasing the current through the wire 

 we can heat it gradually to an intense red or white heat, and we 

 find that as soon as it reaches a dull red, in fact as soon as it is 

 luminous, the air drawn from it has the property of being able to 

 discharge a negatively charged body but not a body positively 

 charged. As the temperature of the wire rises a negative charge 

 is discharged more rapidly, and when we reach a sufficiently high 

 temperature the gas coming from the wire will discharge a body 

 charged either positively or negatively. 



To illustrate these effects we give the following numbers 

 obtained in one experiment. The gas is forced past the incan- 

 descent wire which is substituted for the arc in Fig. 1, and through 

 the tube A past the terminal B, which is charged to a high 

 potential and its rate of leak observed. 



Current heating the Bate of leak of terminal. 



platinum wire. Charged positively. Charged negatively. 



4 - 4 amperes No leak No leak 



5*4 amperes No leak 10 divisions in 60" 



64 amperes No leak 70 divisions in 60" 



7*4 amperes 15 divisions in 60" 90 divisions in 60" 



8'4 amperes 100 divisions in 15" 100 divisions in 15". 



These numbers show that the wire requires to be at a higher 

 temperature in order to discharge positive electricity near it, than 

 is required to discharge negative electricity. Also, at very high 



