250 Dr. R. S. Willows on the Effect of a 



shunted across the coils and the interrupter as in experi- 

 ment (c), its effect on the extra-current at the interruptor 

 must be evidently null. 



Thus we see that, while the rise of current up to the critical 

 value at break is governed by the equation (2), a new start 

 of the current is effected principally by means of the electro- 

 motive force due to the self-induction. It is now very in- 

 teresting to compare Wehnelt's interruptor with the induction- 

 coil. In the induction-coil the suppression of the spark due 

 to the extra-current is indispensable, so that a capacity of a 

 certain requisite value must be inserted across the interruptor. 

 The deficiency of capacity diminishes the oscillations in the 

 primary circuit in consequence of the spark due to the 

 extra-current, and therefore the latter spark must be avoided 

 as much as possible. On the other hand, in Wehnelt's 

 interruptor the spark due to the extra-current is necessary 

 in order to keep up its action. Without such spark, the 

 vapour at the active electrode cannot be got rid of, and 

 consequently sufficient current cannot be established anew. 

 The actions of capacity and self-induction are thus directly 

 opposite to each other in the two cases. 



In conclusion, I wish to acknowledge my indebtedness to 

 Prof. Warburg for encouragement in performing the above 

 experiments. 

 Berlin, December 1900. 



XXI. On the Effect of a Magnetic Field on the Discharge 

 through a Gas. By R. S. Willows, B.A., D.Sc, Trinity 

 College ', Cambridge, 1851 Exhibition Scholar *. 



IT is well known that if a discharge is passing through a 

 tube containing rarefied gas, the effect of a transverse 

 magnetic field is to increase the potential- difference at the 

 terminals and to decrease the current passing through it, 

 while a longitudinal field renders the passage of the discharge 

 easier |. These results are easily explained by the ionic 

 theory. 



While working with tubes in which the pressure varied 

 from *1 mm. to 1 mm. it was noticed that under certain 

 conditions the imposition of a transverse field caused a large 

 increase in the current passing and a decrease in the potential- 

 difference of the terminals, just the opposite of what generally 

 occurs. As this did not appear to be due to any peculiarity 

 of the tube, further experiments were made. The results of 

 these are contained in the following paper. 



A remarkable effect of a longitudinal field has already 



* Communicated by Prof. J. J. Thomson, F.R.S. 

 t See ' Rec. Researches/ p. 105. 



