376 



Prof. A. Schuster. On the 



[June 16, 



sufficient to give an account of those experiments which I consider 

 most conclusive. 



The same vessel was used as in the previous experiment ; the screen 

 was not so long, but left a space of 10 cm. free at the bottom. Fig. 2 

 will explain the arrangement. SS' is the screen, always conducted to 



Fig. 2. 







f 



•A \ 



1 CO 





• B J 







earth ; A and B are the electrodes for the main discharge ; C and D 

 are the auxiliary electrodes, which were of the form of copper cylin- 

 ders, 4 cm. high, and having a diameter of 1 and 1*8 cm. respec- 

 tively. Their distance was 2 cm., and their axis was 2 cm. away from 

 the screen. Of the main electrodes, one was 2 cm. and the other 

 1 cm. away from the screen ; their distance was about 4 cm. The 

 discharge from the battery was always used. Whenever a steady 

 current passed between A and B, an auxiliary battery of Clark cells 

 would send a steady current between C and D, and the lower limit of 

 electromotive force capable of producing a measurable current seemed 

 only to depend on the delicacy of the galvanometer. 



Thus, for instance, at a pressure of about f mm., a current of 0*11 

 ampere was sent from A to B ; and the following currents were 

 obtained when the poles of an auxiliary battery were connected with 

 C and D, the galvanometer being inserted in the auxiliary circuit : — 



40 Clark cells 0*032 micro-ampere.* 



20 „ 0-021 



10 „ 0-014 



1 Leclanche 0*005 „ 



On another day I tried to reduce the electromotive forces still 

 further. The galvanometer, however, had for this purpose to be 

 rendered so astatic that the always changing torsion of the fibre sus- 

 pending the mirror was a source of serious trouble ; for this reason 

 the numbers have not much value in themselves, but there was no 



# 1 mioro-ampere = 1 ampere x 10 -6 . 



