530 



NATURE 



[January 22, 1920 



LETTERS TO THE EDITOR. 



[The Editor does not hold himself responsible for 

 opinions expressed by his correspondents. Neither 

 can he undertake to return, or to correspond with 

 the writers of, rejected manuscripts intended for 

 this or any other part of Nature. No notice is 

 taken of anonymous communications.'] 



Gravitation and Light. 



In a recent letter (Nature, December 25, p. 412) 

 and elsewhere I have expressed doubt as to the 

 security of the inferences regarding the influence of 

 gravitation on the light from distant celestial bodies, 

 vvhich are advanced as tests of the Einstein formula- 

 tion. A closer and less sceptical general scrutiny is 

 possible. The difficulty was to recognise how a 

 theory which professes to supersede an aether with its 

 definite space and time, by concepts purely relativist, 

 could rnanage to effect direct comparison, at a distance 

 and without tracing transmission across the inter- 

 vening space, of the radiations of a molecule at the 

 sun and those of a molecule of the same substance at 

 the earth. This body of doctrine seems, in fact, to 

 consist of two chapters. A blind man could work out 

 the purely relativist theory, which would indeed repre- 

 sent rather closely the process of groping from point 

 to adjacent point in space and time by which he must 

 acquire his own scheme of knowledge. But to com- 

 pare his results with the world of experience a prac- 

 tical astronomer is needed, with very different equip- 

 ment; he relies on the rays of light, in conformity 

 with the optical theory that prescribes their function 

 as messengers across space. 



It thus appears to be necessary to examine directly 

 what changes in the propagation of rays of light would 

 arise under the modified gravitation, "and, if possible, 

 to bring out more explicitly and demonstratively the 

 further postulate that is needed to reconcile them" with 

 the proposed test-relations. The postulate which is 

 sufficient to sustain the optical predictions proves to 

 be this : that all the way to the sun and throughout 

 the solar system the formula for the element of fourfold 

 length by which the nature of the space is deter- 

 mined does not contain explicitly that one co-ordinate 

 which is more especially related to time, but involves 

 only its differential. This is, of course, a reasonable 

 assumption ; but it is of an absolute type regulating 

 the whole space, assumed to be thus settled in advance 

 on the Newtonian plan, not of the relativist tvpe 

 which would profess only to explore it gradually from 

 place to place as it arises. 



But we can analyse further and more definitely. The 

 new theory implies that if this quadratic formula 

 characteristic of the space involved in its product terms 

 the differential of that co-ordinate which stands closest 

 to time, then the velocity of the rays of light in any 

 direction at any place would be different according a"s 

 they are travelling forward or backward. That could 

 only mean that the co-ordinates define at such a 

 locality a frame of reference which is itself in motion. 

 But in motion with reference to what? The relativity 

 of language is doubtless capable of supplying ah 

 answering formula ; but it would onlv be wrapping up 

 in abstractions the simple statement that when at any 

 place the quadratic characteristic of the spacial exten- 

 sion involves the differential of the co-ordinate specially 

 related to time in its product terms, then there is latent 

 in it a specification of its own mode of change at that 

 place with respect to uniform space-time. If no such 

 products are contained, the space is not locally in 

 motion, and we may sav that the frame of reference 

 is fixed in the aether. That is, the fourfold space-time 

 frame in which we set the universe is everywhere de- 

 formed and awry, but it is then nowhere in mpve- 

 NO. 2621, VOL. 104] 



ment relative to light; or, in graphic terms, the co- 

 ordmate system would involve a fourfold curvilinear 

 frame instead of a rectangular one when it is set in a 

 uniform fivefold extension, but it is to be nowhere'in 

 movement when set in that higher auxiliary space. 

 The physical properties of the rays of li"ght caii 

 scarcely be invoked to obtain an astronomical test of 

 results, by providing in their vibrations a universal 

 scale of time, without becoming to the same degree a 

 criterion of the relation to light of the whole con- 

 struction; if they can settle universal time bv optical 

 vibrations, they can equally well be applied "to settle 

 absolute space in each locality. It comes to this, that 

 radiation can be utilised to determine the space and 

 time absolutely. 



This point of view involves no destructive criticism 

 of the substantial and brilliant mathematical theory, 

 which, of course, ought to evolve correctly the con- 

 sequences of the postulates that are put into it. But 

 it does demur to the popular presentation which 

 asserts that space and time and the asther have now 

 been transcended. The outstanding problem, stripped 

 to its essentials, was to find whether gravitation could 

 be brought into line with radiation in this very arrest- 

 ing feature : that the time which is most appropriate 

 by far for its analytical formulation is a changing 

 local time mixed up definitely, though very slighth', 

 with spacial relations. The value of the new theory 

 is that it opened out a way by which this problem 

 could be attacked, while previously no approach was 

 in sight; and, still more important, that it has not 

 improbably Jed to an answer, in the affirmative. This, 

 of course, is a very remarkable consummation, com- 

 parable to Faraday's detection of an influence of mag. 

 netism on light, though more fundamental in that it 

 relates to free space; it must promise substantial 

 advance as regards the fortnulations on which we con- 

 struct our ultimate plan of physical activity, either 

 along its present lines or some other that would repre- 

 sent the result with equal approximation. But beyond 

 that the extreme relativist developments, where "thev 

 are not metaphysical dogmatics, are a verv interesting 

 extrapolation towards the possible or probable phvsical 

 formulation of a universe in which bodies are moving 

 thousands of times as fast as the stars are found to 

 move in our own. 



Reference may be made to forthcoming Proceedings 

 of the Cambridge Philosophical Society and Monthly 

 Notices of the Royal .\stronomical Society. 



Cambridge, January 17. Joseph Larmor. 



The Outlook of British Technical Optics. 4^- 



The symposium and general discussion on "The 

 Microscope : Its Design, Construction, and Applica- 

 tions," held in the rooms of the Royal Society at 

 Burlington House on January 14, under the auspices 

 of the Faraday, Royal Microscopical, Optical, and 

 Photomicrographic Societies, in co-operation with the 

 Optical Committee of the British Science Guild, with 

 Sir Robert Hadfield, president of the Faraday Society, 

 in the chair, was a landmark in the history of British 

 optics. Whether judged by the number, value, 

 and variety of the exhibits and the papers contri- 

 buted, or by the number of people who attended, the 

 symposium was a success. 



At the present time the microscope possesses a 

 unique interest for those concerned with British optical 

 industries. It demands greater technical knowledge 

 and skill in its designer and producer than any other 

 optical instrument, and the demand for it, both 

 actually and potentially, for work of the most far- 

 reaching importance is so great that it may fairlv be 

 said to be the keystone in the arch of an industry 

 which has already been recognised as one of such 



