of Electricity through Rarefied Air. 17 



rpGGiivp n* 



Deflections. 





I Coo LUC U* 



the air. 



Without induc- 



With induc- 



Difference. 



millim. 



tion-coil. 



tion-coil. 





0-24 . 



. 1-75 



1-85 



-f o-io 



0-99 . . 



. 27-55 



50-05 



+ 22-80 



1-55 . . 



. 23-55 



37-15 



+ 13-60 



2-70 . . 



. 23-0 



17-65 



- 5-35 



3-20 . 



. 21-0 



9-80 



-11-20 



4-50 . 



. 4-55 



0-95 



- 3-60 



6-00 . 



. 7-06 



2'25 



- 4-80 



8-65 . 



. . 5-95 



1-85 



- 4-10 



There are some important circumstances to be noticed in 

 this series, to which another series, not reproduced here, was 

 perfectly conformable. Without the induction-coil the deflec- 

 tions increase in magnitude from a pressure of 4*50 to 0*99 

 millim. The value of e—r x l undergoes therefore continuous 

 augmentation from the first limit to the second. Now, be- 

 tween these limits the expression e + ?\l is, according to the 

 series 3 and 4, nearly constant. From this it follows that e 

 increases continuously between the limits in question, while 

 r-J, undergoes continuous diminution. If the pressure descends 

 to 0*24 millim., the sum e + r x l increases notably, according 

 to series 3 and 4. Now, when we increase the resistance of 

 the circuit kcdh (fig. 1), the quantity of electricity that tra- 

 verses the rarefied-air space is diminished, whence it follows, 

 according to series 5, that the deflections are diminished also. 

 It is here that we must seek the reason of the fact that at 0*24 

 millim. pressure of air the deflection amounts to only 1*75 

 scale-division. This series therefore gives the same result as 

 experiment 1. 



On the other hand, the observations taken when an induc- 

 tion-coil was inserted in the circuit at w or at p gave a par- 

 tially unexpected result. The deflections diminished for a 

 pressure between the limits 8' 65 and 2*70 millim. Of the 

 two induction-currents, going in opposite directions, due to 

 the passage of the discharge-current through the coil, that 

 which traverses the spark in the inverse direction of the dis- 

 junction-current, and consequently also traverses the spirals 

 of the galvanometer in a direction opposed to that current, has 

 the property of asserting itself principally or exclusively. The 

 spark, then, acts as a sort of valve; it lets one of the induction- 

 currents pass, but not the other, or at least opposes a greater 

 obstacle to the passage of one of the currents than to the pas- 

 sage of the other. This is perfectly conformable to what I 



Phil. Mag. S. 5. Vol. 15. No. 91. Jan. 1883. C 



