IJO 



NA TURE 



[January 23, 190S 



suggestive. Mr. Pollard and I did not smooth our 

 observations, as has been done in some other cases, because 

 it was desirable to show how much variability is due to 

 local conditions, how much to observational error. 

 Much of what Prof. Brown would attribute to want of 

 homogeneity in our material I know to be due to the 

 changes in external contour, and this is one of the urgent 

 problems of dam construction. Our experimental work 

 shows that very large changes are made by modification of 

 the contour. Is it not possible to reach experimentally 

 better forms for the flank of a dam than are at present 

 in use? 



The cost of experimental work on a suitable scale is 

 prohibitive to the individual experimenter, but it appears 

 to me that something on a really big scale ought to be done 

 by the Institution of Civil Engineers. The problem is 

 a very important one, and I do not believe, notwithstand- 

 ing what I have read to the contrary, that the keen 

 mathematician feels anything but intense awe in the 

 presence of the creative engineer, who has succeeded, not- 

 withstanding the weak-kneed theory provided by men of 

 science, in producing such monumental works as the 

 Assuan and Vyrnwy dams by the force of mechanical 

 appreciation alone. Karl Pearson. 



The Nature of 7 and X-Rays. 



In a letter to Nature of October 31, 1907 (vol. Ixxvi., 

 p. 661), a copy of which arrived here recently, Mr. Barkla 

 criticises a paper of mine which was published in the 

 Philosophical Magazine for October. In that paper I 

 tried to show how closely the properties of 7 and X-rays 

 Were in agreement with the hypothesis that they consisted 

 mainly at least of neutral pairs, and 1 pointed out that 

 even the peculiar polarisation effects which Mr. Barkla 

 had shown to exist might be explained, not unreasonably, 

 as a consequence of the rotatory motion which such pairs 

 would probably possess. I suggested that a pair might be 

 more likely to become entangled with and deflected by an 

 atom revolving in the same plane as itself. 



Mr. Barkla describes an experiment in which he has 

 measured the amount of scattering of X-rays in different 

 directions, and compares his results with those which he 

 expects to obtain as the result of calculations based on each 

 of the two hypotheses in turn, that of the neutral pair 

 and that of the ether pulse. He states the results to be 

 against the former theory. 



But for the result from his calculations in the case 

 of the neutral pair, he makes the assumption that the 

 probable direction of motion of a neutral pair on emergence 

 from an atom with which it has been entangled is in- 

 dependent of its original direction of motion. 



There is no justification for this assumption. It does 

 not even appear to be probable. Consequently, the experi- 

 ment has no value as a critical test. Vet I fully agree 

 with Mr. Barkl.a that the dependence of the amount of 

 secondary radiation upon the angle which its line of motion 

 makes with that of the primary ray is a very proper sub- 

 ject of study ; it might be expected to furnish much-wanted 

 information. 



For this reason I also have been investigating one aspect 

 of this question. With the assistance of Mr. J. P. V 

 Madsen, of this University, I have been comparing the 

 secondary radiations issuing from the two sides of a plate 

 through which y rays are passing. On the ether pulse 

 theory there should be complete symmetry, provided that 

 the rays have not been appreciably absorbed on their way 

 through. Secondary radiation, whether material or not, 

 originated in an atom by a passing pulse, is just as likely 

 to go forwards as backwards. This is, indeed, always 

 assumed by writers on the ether pulse theory, e.g. by 

 Mr. Barkla himself in the letter already referred to. 



On the other hand, if the 7 rays are material, it is quite 

 possible, though not necessary, that the secondary radia- 

 tions on the two sides of the plate should be different. 



As a matter of fact, there is the most remarkable want 

 of symmetry, and this is fatal to the ether pulse theory 

 of the 7 rays. Moreover, all our experiments so far show 

 that, on the whole, the kathode radiations from a given 

 stratum of matter traversed by 7 rays possess momentum 

 in the original direction of motion of the rays, and this 

 shows that the rays are material. 



NO. IQQS, VOL. 77] 



The experiments are very simple, and are' not wholly 

 new. The secondary kathode radiations due to a stream 

 of rays impinging on a plate have been studied by many 

 observers, who all concur in the statement that they 

 increase with the atomic weight of the material of the 

 radiating plate. For example, the figures for Pb, Al, and 

 C are about as 100 to 30 to 15. In fact, they follow almost 

 the same law as in the case of 3 particles. The reason for 

 this will appear presently. The secondary kathode radia- 

 tions that appear where 7 rays emerge from a plate have 

 been less studied, but Eve has shown that they consist 

 largely of kathode particles, and Dawes [Phys. Rev., xx., 

 p. 182) that they do not appear to follow the same law 

 as what may be called the " incidence " radiations. 

 Further, Wigger has made the very important observation 

 (" Jahrbuch der Radioaktivitat," Bd. ii., pp. 42S-430) that 

 in certain circumstances the 7 rays issuing from a plate 

 of Al which they have traversed make viore secondary rays 

 than when they issue from Pb. 



All these facts may be conveniently studied together. 

 Let an ionisation chamber be made of cylindrical form with 

 plane ends, and let a pencil of 7 rays be directed along the 

 axis. Let the rays first pass through a cm. or so of 

 lead. Let them then pass through a pair of plates which 

 can be inverted ; a convenient pair may be made of a 

 lead plate i mm. and a carbon plate i cm. thick ; these 

 are to form one end of the chamber. It will then be 

 found that there is more current through the chamber 

 when the C plate is next to it, and the " emergence " 

 secondary rays are produced in the carbon, than when the 

 Pb plate is next the chamber. But if the plate closing 

 the other end of the chamber be at one time Pb, at another 

 time C, the reverse effect occurs. At this plate " in- 

 cidence " secondary rays are produced by the same pencil 

 of 7 rays, and these are greater when the plate which is 

 struck is Pb than when it is C. The differences are of 

 the order of 10 per cent., 20 per cent., up to 60 per cent., 

 depending on the circumstances of the experiment. The 

 materials and the form of chamber may be largely 

 varied, but the same want of symmetry is always there. 



It appears that the " emergence " radiations from the 

 plate through which the 7 rays have entered are more 

 important than the " incidence " radiations from the other 

 plate. The latter serves mainly as a reflector or Scatterer 

 of the rays from the former, and this is the reason why 

 there is less current when it is formed of a material of 

 smaller atomic weight, following the usual rule for j3 rays. 

 This effect is a general one, and it serves to explain why 

 all observers have found less kathode radiation due to 

 7 rays from Al or C than from Pb, when actually the 

 rays produce more from the former than the latter when 

 they have been first sifted by a cm. or so of lead. A 

 stream of 7 rays contains /3 particles derived from the 

 7 rays by the influence of the last substance traversed. 

 It is the scattered portion of these which constitutes the 

 main portion of the secondary radiations due to 7 rays, and 

 the reason why the incidence radiations run parallel to 

 those of /3 rays is obvious. Nevertheless, there are small 

 variations due to secondary rays formed in the material 

 of the plate itself, the quantity of which is influenced by 

 the nature, not only of this material, but of the screens 

 through which the rays have previously passed. This is 

 because the 7 rays are heterogeneous, as first shown by 

 Kleeman. It is when the rays have passed through some 

 thickness of lead that they are acted on with greater effect 

 by Al or C than by lead. The quantity of kathode radia- 

 tion set free in the radiator itself depends on the quality 

 of the rays as well as on the radiating material. The 

 particles at first move mainly, perhaps entirely, in the 

 original direction of motion of the 7 rays, but are sub- 

 sequently scattered, and contribute to some extent to the 

 " incidence " secondary radiation. But the principal por- 

 tion of the incidence radiation is due to particles which 

 were in the stream of 7 rays before it struck the plate. 



It would make this letter too long if I were to discuss 

 these results with any fulness, or to show their relations 

 to results already obtained. I hope to publish a fuller 

 account in a short" time. Meanwhile, I will point out that 

 the experimental proof of the material nature of the 7 rays 

 carries with it, almost surely, a corresponding proof as 

 reg-ards (he X-rays. The ooints of similnritv are too 

 numeroi- fcr it to be otherwise. Only, as I have said in 



