356 Separator and Shunt for Currents of High Tension. 



mental purposes; so that, although the last-mentioned arrange- 

 ment might serve for a display, it can hardly be said to give 

 much promise of practical utility. 



We next devised another plan whereby both sets of currents 

 can be turned to account in the same tube with much advan- 

 tage. If instead of two tubes we take a single tube and fur- 

 nish it with two sets of terminals, then, if we connect one pair 

 of terminals of the separator with one pair of terminals of the 

 tube, and the other pair of the separator with the other pair of 

 tube, in such a way that both sets of currents enter the tube 

 at the same end, we shall have a double set of discharges in 

 the same direction through the tube. 



It was, however, further found that it was uot necessary to 

 have four terminals in the tube ; for, taking advantage of the 

 difficulty in effecting a discharge when the negative terminal 

 is small, we fixed two very small terminals near one end of 

 the tube and one large one at the other. We then connected 

 the two small terminals directly with the two sides of the 

 machine respectively (that is, before the currents entered the 

 separator), and the one large terminal with the two sides of the 

 machine after the currents had been led through the two 

 branches of the separator respectively. This effected the 

 object desired ; for, in the first place, there was no tendency 

 in the discharge to pass from one small terminal to the other 

 on account of the principle above quoted; secondly, there was 

 no tendency in the current to become short-circuited at the 

 single large terminal, on account of the difficulty in the pas- 

 sage of the current backward through either branch of the 

 separator. Consequently every current issuing from the 

 machine was obliged to pass to one or other of the small ter- 

 minals, and thence through the tube to the large terminal at 

 the other end. The scheme of the connexions to the various 

 parts of the apparatus will, however, be best understood by 

 reference to the diagram (Plate IX. fig. 2). 



The arrangement may therefore be said to have completely 

 solved the problem of producing unidirectional discharges 

 through a tube from an alternating machine, and at the same 

 time utilizing all the currents from the machine. The method 

 has also the advantage of making the number of currents 

 discharged through the tube in a given time double that 

 which would have been discharged if one set of currents had 

 been merely diverted as in the arrangement first described. 



But beside shunting off, or diverting, one set of currents, it 

 was often desirable to reduce the strength of the whole to suit 

 the condition of the tube or the circumstances of the experi- 

 ment. The readiest method of effecting this appeared to be 



