596 Mr. A. M. Tyndall on the 



has also been obtained by Lonsdale* for their existence at 

 atmospheric pressure. It has further been suggested by 

 Sir J. J. Thomson f that the emission of these bodies accom- 

 panies the process of ionization, so that it is not unreasonable 

 to suppose that such are emitted from the glow region near a 

 point. If their initial velocity is high, they may travel some 

 distance from the point before their kinetic energy falls to 

 that of the molecules surrounding them. 



ISTow suppose that they are emitted from N and that the 

 distance y is a measure of their range : then they will have no 

 effect on a point P when it is placed at a distance y greater 

 than y . But when P is brought within the distance y , 

 doublets possessing considerable energy will strike its surface. 

 By this bombardment energy will be communicated to P in 

 an amount which will increase as y is made smaller, and this 

 supply of energy may be available to aid the process of 

 ionization at P, so that the nearer that P approaches N, 

 the smaller will be the field necessary to start the discharge 

 at P. 



The supply of doublets, however, must be kept up by N, 

 hence the size of P will not affect the value of y . Also the 

 range of the doublets will be independent of N if it is 

 assumed that the initial velocity of expulsion from N is 

 constant. Some of the doublets in the distance y will either 

 ionize the gas or will themselves break up, so that z receives 

 an explanation on the lines suggested in (1). It is consistent 

 with this view that z is independent of the size of the dis- 

 charging point, and is practically the same for positive and 

 negative discharge in air. 



Now z is much less than y . It is, however, really the 

 value of z from to A in Curve 5 that should be compared 

 with y : this value is about 0*8 cm. in air. Also it is pos- 

 sible that z may be too low because the wind pressure 

 method may not be sensitive enough to detect the small 

 number of ions occurring at greater distances. 



In negative discharge in very pure hydrogen r is many 

 times greater than in air, but for various reasons, detailed 

 below, the growing ion view gives in this particular case a 

 better explanation. 



It seems necessary to suppose that these doublets are pro- 

 duced only during the ionization of molecules by means of 

 positive ions. If, for instance, doublets were produced during 

 ionization by negatives they would take the place of the 

 positives in being a fresh source of negative ions, and the 



* Lonsdale, Phil. Mag. [6] xx. p. 464 (1910). 

 t Thomson, Phil. Maj. [6] xviii. p. 821 (1909). 



