Supplement to '■'Nature'' Augtist 4, 1923 



189 



It is unnecessary to point out in the first instance 

 that hght is now known to exert pressure^ and there- 

 fore to convey momentum. An advancing wave-front 

 possesses momentum^ which it can transmit to any 

 obstacle which either reflects or absorbs it. If reflected^ 

 the pressure it exerts is double what it exerts when 

 absorbed ; all quite in accordance with common sense. 

 But I rather want to concentrate attention on the 

 state of things when the wave-front is advancing — 

 it may be for hundreds of years — through so-called 

 empty space. It carries with it a pressure equal to 

 the energy per unit volume. If the Third Law of 

 ^lotion is true without exception^ — and it is surely 

 politic to assume the truth of that law until it is 

 negatived, — there must be a longitudinal stress in 

 that stream of light, with a reaction on the source 

 at one end and on the advancing wave-front on the 

 other. 



The source is always something material. Light 

 can only emanate from an accelerated — that is, from 

 a revolving or vibrating — electron. Hence, at that 

 end the reaction has a material basis, in accordance 

 with the customary experience that a hne of stress 

 must stretch from one piece of matter to another. 

 But what happens at the other end } When it 

 encounters matter, the reaction is exerted on that 

 matter, and everything is plain sailing. But while 

 it is advancing in free ether, what is it that sustains 

 the reaction ? We can only answer, the wave-front. 

 The wave-front cannot sustain it statically. It can 

 only do so by advancing at the speed of light. But 

 it is remarkable and worthy of note that in this 

 particular the advancing wave-front simulates one of 

 the properties of matter, namely, the power of sustain- 

 ing stress. 



Now to me this is very suggestive. We do not 

 know what precisely is the kind of motion occurring 

 in the associated electric and magnetic vectors which 

 are travelling with the speed of light. We do not 

 know the kind of motion associated, more statically, 

 with an electron. But the guess is almost forced 

 upon us that possibly these two kinds of motion are 

 not entirely distinct. We could not say that perhaps 

 they are one, and not two ; for there are certainly 

 differences between them. One must advance ; the 

 other may stay still. But is it possible to regard one 

 as a consequence, or as a generator, of the other ? 



The electron generates light. 



Does light generate an electron ? 



(I am using the term " Light " in a very general 

 sense, not limiting it to the physiological kind which 

 excites the sense of vision, but including X-rays, and 

 Jill other forms of short wave radiation.) 



What do we know about the effect of this kind of 

 radiation upon matter ? We know that it can produce 

 the irregular movements that we call heat, and also 

 that it can stimulate chemical action. But the 

 discovery of photoelectricity shows us that it may 

 do more. It may fling out an electron, with a surprising 

 amount of energy, dependent upon the frequency, 

 that is, upon the wave-length, of the incident radiation. 

 Tills is a hint not to be ignored. Nor is it ignored ; 

 and there must have been many speculations as to the 

 kind of way in which it achieves this result. One 

 would naturally suppose at first that it must do it 



by means of resonance, that is, by the accumulation 

 of properly timed impulses, until an explosion occurs. 

 But the evidence is, on the whole, rather against a 

 resonance view ; because the result seems almost 

 independent of the intensity of the incident radiation, 

 and to depend only on its wave-length. Nor does 

 it seem as if a great length of radiation was necessary 

 in order to produce the result : though this is a matter 

 which requires further and more conclusive experiment. 

 If a beam of light is interrupted and cut up into small 

 sections — as might be done by a narrow slit in a very 

 rapidly revolving disc, — would this intermittent light 

 be equally effective ? For if it is equally effective, 

 the fact would tend against the continuous accumula- 

 tion of a small synchronous disturbance. 



I believe that some experiments have been made 

 in this direction, and that the answer — so far as it 

 goes — is that intermittent is as effective as continuous 

 illumination, and feeble light as efficient as strong. 

 The energy falling upon a minute surface in a beam 

 of diffuse light is insufficient to account for the 

 energy of the resulting effect, unless it is a trigger 

 effect. 



But this rather wants pressing to extremes. To 

 cut up a beam of light into really short portions is 

 not very easy. If a radial slit a millimetre in width 

 is made in a disc a metre in diameter, revolving a 

 hundred times a second, light sent through it is diluted 

 and cut up into sections ; but the length of each 

 section is still about a mile, and accordingly would 

 contain more than a thousand million waves, — which 

 is amply sufficient for resonance. 



However, the evidence so far is supposed to negative 

 the resonance idea : so much so that it has been 

 supposed that the wave-front is not a uniform surface, 

 but a speckled one ; that it is discontinuous ; and that 

 the amount of energy concentrated in one of the 

 specks may be vastly greater than would be reckoned 

 on the diffuse or continuous theory. 



The idea of a speckled wave-front would have 

 seemed to our scientific ancestors rather wild ; though 

 it must be remembered that Newton, with his Corpus- 

 cular Theory, was temporarily satisfied with something 

 very like it. Nevertheless, the Corpuscular Theory 

 had to be abandoned because of the artificial way in 

 which it explained polarisation, and because it seemed 

 to require that light should travel quicker inside 

 matter than outside, instead of slower — as Foucault 

 proved it to do, — and because there are real difficulties 

 in explaining interference and diffraction, unless the 

 wave-front is continuous. 



However, it does not do to turn down a theory too 

 readily and prematurely, merely because we encounter 

 a few difficulties. No hypothesis is wild which has 

 attracted the serious attention of J. J. Thomson, 

 and other brilliant physicists, including — as I think 

 we must — even Faraday ; as evidenced by his 

 " Thoughts on Ray Vibrations " (" Experimental 

 Researches," vol. iii. p. 447). 



Moreover, though these ideas, as we perceive them 

 at present, may not be able to substantiate themselves, 

 yet they are the outcome of observed facts ; and it 

 may yet be found that, in a modified and revolutionised 

 form, they may contain elements of truth at present 

 unsuspected. 



