444 Mr. J. Henry on the Deflexion by Magnetic 



(2) The Rate of Recombination of the Ionized Gas. 



The recombination of the gas ionized by a discharge has 

 much the same effect as diffasiou in destroying the conduc- 

 tivity of the gas between successive discharges, rapid recom- 

 bination acting like rapid diffusion. 



(3) The Velocity of the Ions in the Gas. 



The velocity of the ions in the gas under an electromotive 

 force has an effect similar to (1) and (2), since the ions will 

 be pulled towards the electrodes between successive discharges, 

 thus destroying the conductivity of the path through the gas ; 

 the greater the velocity of the ions the less would be the 

 conductivity of the ionized gas left by a discharge for the 

 following one. 



For high pressures of the gas in the discharge-tube the 

 effect of the velocity of the ions should be more marked, when 

 an induction-coil is used to send the discharge. In this 

 case the coil will only send a discharge for a break of the 

 primary, while for the make there is an electromotive force 

 acting on the gas between the electrodes pulling in the ions, 

 though no discharge passes to renew the supply. 



(4) The Duration of the Discharge. 



The time given for causes (1), (2), and (3) to destroy the 

 conductivity of the gas for a given frequency of discharge, is 

 diminished by the length of time each discharge lasts, so that 

 the greater the duration of each discharge the less will be the 

 effect of (1), (2), and (3) in destroying the conductivity of 

 the gas between successive discharges. For very low pressures 

 of the gas in the discharge-tube, the time each discharge lasts, 

 as shown by a rotating mirror, is a big part of the interval 

 between two discharges produced by an induction-coil. 



(5) The Conductivity of the Gas after a Discharge has passed, 

 through it, relative to that of the same Gas before the 

 Discharge. 



If a gas suffers little change from the passage of a dis- 

 charge through it, and if its resistance to the first discharge 

 is little greater than it is to succeeding discharges, then the 

 path of the discharge should never be much deflected from the 

 straight line between the electrodes, as the length of the path 

 of least resistance for the discharge through the gas cannot 

 in this case be much increased. 



In hydrogen, for example, the potential-difference required 

 to maintain a discharge between a pair of electrodes is much 



