52 



hypothesis that the energies of the primary electron, the 

 bundle, and the secondary electron are all equal. The whole 

 of the energy of the cathode particle in the X-ray tube is 

 converted into one energy bundle. This darts away from the 

 anticathode, and sooner or later causes the ejection of an 

 electron from some atom which it traverses, handing over 

 to the electron the whole of its own store of energy. Replace 

 the bundle of energy by a neutral pair, and the whole affair 

 seems simple enough. But surely the complications of the 

 ether structure increase the more closely we examine the pro- 

 cess under which rapidly moving electrons in the X-ray tube 

 disappear, and similar electrons, moving at the same rate, ap- 

 pear elsewhere, if we are to consider that the only links be- 

 tween them are little bundles of energy moving with the 

 speed of light. 



It might be said, perhaps, that one bundle contains the 

 energy of several arrested electrons ; but in that case we 

 should have bundles of all sizes and secondary electrons of 

 all speeds : or that several bundles might pile up their ener- 

 gies in one atom until there was enough for the ejection of 

 one secondary electron ; but then we should return to the dif- 

 ficulty of explaining why the speed is independent of the na- 

 ture of the atom. 



A cathode particle cannot give all its energy to a pulse 

 unless its arrest is brought about in a very sudden and spe- 

 cial way. The thickness of the pulse must not be greater 

 than the diameter of the electron or corpuscle ( 'Cond. of 

 Elect.," p. 660). If the pulse is thicker than this, only a pro- 

 portional fraction of the energy of the cathode particle can 

 be converted into the energy of the pulse. Now it is gene- 

 rally believed that the phenomena of the X-rays require a 

 pulse many thousands of times as thick as the diameter of 

 an electron. It does not seem possible to reconcile these op- 

 posite requirements. 



The bundle must be excessively small. If it is larger 

 than an atom, or even than an electron, the whole of its 

 energy cannot be given up to one electron on which it im- 

 pinges. There would not be time for the energy to move 

 in from the outskirts of the bundle to the place where it is 

 being transformed into the motion of the electron. It must 

 not expand or contract its borders as it moves, or else its effect 

 will vary as it travels. 



The difficulties of this theory are exactly those which 

 would naturally arise in the attempt to transfer the proper- 

 ties of a material particle to an immaterial disturbance. 



Let us now consider the third form of hypothesis, ac- 

 cording to which both the energy and the material of the 



