426 Dr. Tyndall and Mr. Hughes on Cathode 



of low volatility, the phenomenon of disintegration is much 

 simplified and can be readily controlled from the very start 

 of discharge. 



It has generally been agreed that the disintegration is 

 •caused by positive rays. Thus Stark regards a particle of 

 disintegrated cathode as a secondary " atom-ray/' and making- 

 certain arbitrary assumptions he has deduced an expression 

 ;f or the velocity of expulsion of such rays. One would expect 

 that the rate of disintegration of a cathode would depend on 

 the number, the energy, and the specific charges of the par- 

 ticles striking it. Now the positive rays are complex in 

 these respects. J. J. Thomson classes under positive rays 



(a) rays of atomic or molecular size with simple or multiple 

 positive charges and of varying velocity acquired by falling 

 through the whole or part of the Crookes dark space ; 



(b) similar rays, but possessing a constant high velocity un- 

 influenced by the cathode-fall ; (c) uncharged rays also of 

 high velocity. There is no means at present of determining 

 what individual part these different forms play in the pro- 

 duction of disintegration, and a theoretical analysis of the 

 curves obtained would therefore be premature. 



There is one point, however, which may be noted. In 

 previous discussions on the effect of current density, it has 

 generally been tacitly assumed (1) that all or a constant 

 fraction of the positive rays arriving at the cathode are 

 •effective in producing disintegration, and (2) that they carry 

 a constant proportion of the current in the dark space. On 

 the former point there is no information. Some experi- 

 ments by Hodgson and Mainstone * on the heating effect of 

 the " normal " discharge at the cathode lend support to the 

 latter assumption; but it is doubtful whether their conclusions 

 •can be applied to the above results, because in these the 

 •cathode-falls were in general much higher and the cathodes 

 used were very small. Under these conditions much higher 

 temperatures would be obtained, and it might be expected 

 that an increase in cathode-fall or current might cause an 

 increase in the proportion of current carried by negative 

 ions in the dark space. This may explain why the propor- 

 tionality between current density and rate of disintegration 

 previously observed at small values of current is not main- 

 tained in the general case. 



It is also interesting to note in this connexion that the 



marked flattening of the disintegration-cathode-fall curves 



which sets in at high cathode-falls occurs more readily in 



Jhydrogen than in other gases ; it is well known that the 



* Phil. Mag. xxvi. p. 411 (1913). 



