i6 



NATURE 



[September 4, 1919 



at all ; you cannot think of such a body on the earth, 

 but you can think of one under no resultant force, 

 i.e. under balanced forces. Such a body moves by 

 reason of its inertia alone. It is in equilibrium ; it 

 is not at rest. 



But we have no sense of straightforward locomo- 

 tion, and not the slightest clue to either the magni- 

 tude or direction of our motion through space. We 

 can ascertain approximately how the sun is moving 

 with reference to our system or cosmos of stars, but 

 we do not know at what rate that system is itself 

 moving.. For all we know, it may be moving very 

 fast, hundreds of miles per second. 



We have a sense of acceleration, however; we 

 experience it in a lift as it begins to descend; and 

 if the sensation is repeated often enough, as on a 

 rough sea, the result is unpleasant. We have also a 

 sense of rotation ; we can tell when our vehicle — 

 say a Tube train — turns a corner in the dark. Most 

 animals appear to have a sense of rotation, apparently 

 located in the ear. But we have no sense of direct 

 translation ; and we have so far failed to devise any 

 instrumental means for detecting our motion through 

 the aether of space. 



The failure is not for lack of trying. Many experi- 

 ments have been tried, but there is always some 

 compensating effect; so we get no answer to the 

 question : \t what rate and in what direction .ire we 

 moving? The best known experiment is that of 

 Michelson and Morley, the result of which seems to 

 assert that the aether clings to the earth, or that the 

 earth is not moving through any kind of substance. 

 But Fizeau's classical experiment showed that a 

 transparent body carried with it none of the internal 

 EBther of space ; and experiments made by myself " at 

 Liverpool in the nineties of last century showed that 

 a rapidlv moving opaque body carries no external 

 aether with it, that there is no perceptible viscous 

 drag or cling between matter and aether, and accord- 

 ingly demonstrates that stagnation or absence of 

 relative aether drift past the earth is not a reasonable 

 explanation of Michelson 's negative result. 



The two experiments together, in fact, ought to be 

 taken as establishing the reality of the most interest- 

 ing of all the compensating effects yet discovered, 

 viz. the FitzGerald-Lorentz contraction of all matter 

 in motion, which the electrical theory of cohesion 

 renders so extremely probable. It only amounts to 

 a 3-in. shrinkage in the whole diameter of the earth 

 in the direction of motion ; but it is enough. This 

 slight contraction or change of shape in moving 

 bodies I regard as the definite and interesting com- 

 pensating effect in this ca.se. Incidentally, moreover, 

 it establishes the electrical, i.e. the chemical, nature 

 of cohesion. For, given that cohesion is a residual 

 chemical affinity — due to the outstanding attraction 

 of molecules composed of neutral groups ot equal and 

 opposite electric charges, brought so near together 

 that the attraction between molecules is no longer 

 averaged to zero' — then, on orthodox Maxwellian 

 electric theory, a diminution of this force due to 

 lateral motion is inevitable. And the resulting lateral 

 expansion or longitudinal contraction, or both, is of 

 the right order of magnitude. So this acts as a 

 previously quite unsuspected compensating effect, 

 which exactly neutralises the drift effect otherwise to 

 be anticipated. Thus, by superposition of two posi- 

 tive consequences of drift, the Michelson experiment, 

 like every other yet made, declines to indicate that 

 there is any drift at all. 



Hence, after many such negative results, it seems 

 to become hopeless to inquire experimentally as to 



2 See Phil. Tran«., vol. clxxxiv. (1893), pp. 727-804, and vol. clxxxix. 

 1807), pp. 140-^6. 

 ^ See, for instanre, my book on electron*!, chap. xvi. 



NO. 2601, VOL. 104] 



our motion through aether, unless, indeed, gravitation 

 were exempt from the otherwise universal compensa- 

 tion.. In that case the electrical theory of matter 

 applied to the motion of planets might yield a residual 

 result. But my recent inquiry into this problem has 

 suggested that gravitation, too, is in the conspiracy,' 

 and in that case there is some ground for the con- 

 tention of the extreme Relativists, not only that we 

 do not know our motion — with which everyone agrees 

 — but also that we never shall know it ; and, in fact, 

 that motion of matter through aether is a phrase 

 without meaning. 



I hope wo shall not too readily shut the door on 

 further attempts in this direction ; and as a conserva- 

 tive physicist I may be allowed to lament the extra- 

 ordinary complexity introduced into physics and into 

 natural philosophy by the principle of relativity, as 

 so remarkably and powerfully developed by the mathe- 

 matical genius of Einstein, with complication even 

 of our fundamental ideas of space and time. The 

 complications do not commend themselves to all of 

 us, and I for one should be glad to return to the 

 pristine simplicity of Newtonian dynamics, modified, 

 of course, by the electrical theory of matter; ad- 

 mitting the FitzGerald-Lorentz contraction, and 

 admitting also the variation of effective inertia with 

 speed. These things do not destroy, but supplement, 

 Newtonian dynamics. They generalise it in a legiti- 

 mate and intelligible manner. Such complications as 

 these are clearly in accordance with truth, and are 

 to be welcomed; but the complicated theory of 

 gravitation created this century by Einstein, and 

 developed by his successors, and the consequent over- 

 hauling of space and time relations, do not at present 

 commend themselves to me, or, I think, to others 

 of what I suppose must be called the older school. 



Meanwhile, the full-blown theory has the courage 

 of its conviction and has predicted a definite result, 

 viz. the deflection of a ray of light by the sun's limb, 

 equal to 175 seconds of arc. The prediction is going 

 to be tested during the solar eclipse of May 29 this 

 vear, between Brazil and the Gulf of Guinea, l^et 

 the issue be clearly understood. If a star-ray grazing 

 the sun is deflected f second it will mean only that 

 light has weight, that the wave-front not only simu- 

 lates the properties of matter by carrying momentum 

 — as we know it does from the investigations of 

 Nichols and Hull, Poynting and Barlow, and others — 

 but that it is even .subject to gravity. For this would 

 be the angle between the asymptotes of a cometary 

 orbit when the comet is moving with the speed of 

 light and passing close to the sun.' But the principle 

 of relativity — through the refractive or converging 

 influence of a strong divei-gent gravitational field — 

 demands a greater deflection than this, more than 

 twice as great. .So there are three alternate deflections 

 Ix'fore us, to be settled by observation : — 175 sec. ; 

 0-75 sec. ; and zero. Let us hope that the result of 

 this or of some other eclipse-opportunity may be 

 definite enough to discriminate clearly and quantita- 

 tivelv between these three alternative values, any 

 one of which should be equally welcome to any lover 

 of truth. 



If the first answer is given decisively, it will be a 

 conspicuous triumph for the theory of relativity, 

 and will for a time be hailed as a death-blow to the 

 aether. I claim beforehand that such a contention is 

 illegitimate, that the reality of the aether of space 

 depends on other things, and that the establishment 

 of the principle of relativity leaves it as real as before; 

 though trulv it become.s even less accessible, less 



■* See the Phil. Mig. for August, igi?, and February, IQ18, pp. 145, 155 

 and 1^6. 

 » See, for instance, my paper in the /'*//. Mk/;. for August, 1917, p. 93. 



