TRANSACTIONS OF SECTION A. 541 



or light as usually conceived. The point to be considered now is the classification 

 of X and 7 rays. 



The one certain property of the X or 7 ray is its power of exciting a £-ray by en- 

 counter with an atom. The speed of the /8-ray depends on the qualit}' of the X-ray, 

 but not on the nature of the atom, and its direction continues more or less the direction 

 of the X-ray, more in the case of the penetrating 7-ray, less in the case of the ' softer ' 

 X-rays. Since /3-rays produce ionisation, phosphorescence, and photographic action 

 it may be true, and experiment actually proves, that these accompaniments of X-ray 

 absorption are really due to the 18-rays, the X-rays having no direct effects of this 

 kind. The ' absorption coefficient ' of a substance is simply a measure of its power 

 of prompting the X-rays to produce /3-rays. 



The energy of the /9-ray cannot come from the atom ; for each atom would then 

 have its own speed of ejection which, as well as the direction of projection, would be 

 independent of the properties and conditions of the X-ray. Also there is no evidence 

 of the existence of such induced radio-activity, implying the release of atomic energy. 



Nor can the energy of the /8-ray be the result of the accumulation in the atom 

 of energy extracted from many X-rays. The arguments against the case just con- 

 sidered hold here also with little change : and it could hardly be supposed that the 

 accession of the last infinitesimal amount of energy required to fill the store in the 

 atom and provide for the energy of the /?-ray would determine so effectively the 

 direction in which the /8-ray is ejected. 



We must, therefore, suppose that one X-ray provides the energy for one /3-ray; 

 similarly that in the X-ray bulb, one /3-ray excites one X-ray. I pointed out two years 

 ago ' that this was strongly supported by the near equality in speed of the /8 or cathode 

 ray which produces the X-ray in the X-ray bulb to the speed of the /3-ray which that 

 X-ray finally produces. The figures were only approximately, though sufficiently, 

 known at that time : the recent determinations of Whiddington have greatly 

 strengthened the argument. It is clear that little energy is lost in the interchange of 

 form, /3 to X ray and back again : and further, since the speed of the secondary 

 /3-ray is independent of the distance which the X-ray has travelled, the X-ray cannot 

 diffuse its energy as it goes. That is to say, it is a ' corpuscle.' Moreover, it cannot 

 spend energy on ionisation, a conclusion which I have verified. 3 The energy of an 

 X corpuscle is given by the energy of the /3-ray which it produces : and this definition 

 is complete so far as is known. It is preferable to a definition by penetrating power, 

 since the latter must have reference to some particular substance. 



Barkla has shown that many substances emit homogeneous X-rays of characteristic 

 penetrating power when irradiated by X-rays of greater penetrating power. In the 

 language of the corpuscular theory, an X corpuscle of definite energy is frequently 

 emitted by a substance which is under bombardment by X corpuscles of greater 

 energy. As far as the energy relations go, this is what the corpuscular theory would 

 lead us to expect. Whiddington has shown 3 that X-rays cannot excite the charac- 

 teristic rays of any substance unless they have themselves been excited by cathode 

 rays of energy exceeding a certain limit. This follows from the corpuscular theory. 

 The requisite energy is that of the characteristic X corpuscle of that substance. The 

 recent demonstration by C. T. R. Wilson of the tracks of ionising agents in a gas, 

 tracks which are rendered visible to the eye by the condensation of water vapour 

 upon them, are also illustrations of the theory. 4 



The corpuscular form of the X and 7 ray, and its energy relations to the /3-ray 

 which is its origin, and the /3-ray which is its conclusion are the principal things which 

 any theory must account for, and any model must illustrate. For this reason the 

 spreading pulse of Stokes fails : and so also does the kink in the tube of force of 

 J. J. Thomson. The former diffuses its energy over a broadening surface, the latter 

 over a lengthening line. It is true that certain phenomena suggest the existence of 

 links between X-rays and light, such as the so-called polarisation of the former, and 

 the photo-electric properties of the latter. But the nature of these links is not clear 

 now, though it may be revealed more clearly in the future. The ' corpuscular theory ' 

 of X-rays is really an induction from experimental facts. The ' neutral- pair ' theory 

 which I have described at various times is at any rate a simple working model. But 

 the ' ether-pulse theory ' is at present little more than an aspiration. 



' Presidential Address to the Australasian Association for the Advancement of 

 Science, Brisbane, January 1900. - Proc. Roy. Soc, 85, p. 349. 



3 Ibid., 85, p. 323. • Ibid., 85, p. 285. 



