Febru^RV ij, 192 1 J 



NATURE 



809 



that it is impossible to say definitely whether it 

 has been observed or not. But it is not so well 

 known that the assumption that would lead to no 

 >pectral shift leads to the result that the wave- 

 length on emission of light from a oarticular type 

 of atom is a function of position ; thus the aban- 



donment of ds as the fundamental measured quan- 

 tity would not make it possible to carry both 

 length and time standards about unaltered. 

 Probably the difficulties arising from the hypo- 

 thesis that ds does not play two parts are so great 

 as to render it quite unplausible. 



The Metaphysical Aspects of Relativity. 

 By Pkof. H. WiLDON Carr. 



'r*HERli is a possible misconception in the ap- 

 ^ plication of the term "metaphysical" to the 

 new principle of relativity which it is advisable 

 to clear up. In the great era of the triumphant 

 advance of the positive sciences, which began 

 about the end of the first third of the nineteenth 

 century, metaphysics was decried as the main 

 obstacle to scientific progress. Following the lead 

 of .\ugustc Comto, the workers in the sciences 

 held it up to scorn as obscurantism. The derision 

 and reproach which were then poured on it have 

 clung to it ever since. There are many to-day 

 who acknowledge, indeed, that metaphysics must 

 be assigned a place in the hierarchy of the 

 .■-ciences, but interpret the Aristotelian defini- 

 tion, "that which follows or comes after physics," 

 as indicating a dark realm of the yet unknown, 

 or even of the unknowable, which surrounds the 

 I lear zone of positive knowledge, into which we 

 may peer, but will discern nothing. The objects 

 of metaphysics — the soul, the cosmos, the deity 



are in this view vain imaginings, not objects of 

 which there can be knowledge in the scientific 

 meaning — that is, objects amenable to the ex- 

 p<'rimcntal method. Such a view simply ignores 

 the scientific tradition. Modern science is the 

 result of the formulation and adoption of the ex- 

 perimental method, but the experimental method 

 is not self-evident or inherently rational ; it de- 

 pends on a metaphysical concept, and its ration- 

 ality can be established only by metaphysical 

 principles. To contrast, then, the experimental 

 method with the principles on which it depends, 

 to describe one as the realm of science and 

 the other as the realm of ignorance or unknow- 

 ibility, is from any philosophic point of view 

 -lultifying, and, in the literal sense, absurd. 



What has made it possible to consider meta- 

 physics as an unreal science, or as a realm of 

 unreal fancy, is the peculiar position in regard to 

 the natural sciences in which the purely mathe- 

 matical sciences stand. Mathematics does not use 

 the experimental method, and in the hierarchy of 

 the sciences mathematics seems sufficient of itself 

 for the foundation and support of the whole super- 

 structure. But mathematics is only an abstract 

 s<-icnce of quantity ; its concepts lack the one 

 I'sscntial character which experimental science 

 • alls for conrrct<-ness - and this metaphysics 

 ilonr can supply. 



The modern era of philosophy from Descartes 

 mwards has t)een dominated by the insistence of 

 ih( scientific problem — that is. the problem of 



N'O. 267 7. VOI . 106] 



the ultimate nature of tho reality we study in 

 physical science by the experimental method. 

 This interest in the nature of scientific reality 

 replaces the main interest of the philosophy of 

 the mediaeval period, whicl, was concerned with 

 the origin and destiny of the human soul, and, 

 more generally, with the relation of man to God. 

 If modern philosophy may be said to join hands 

 with the ancient philosophy of Greece, it is not 

 in the identity of it.< interest ; for, though the 

 (Jreeks were mathematicians, they had no con- 

 ception of the experimental method as we practise 

 it, and it is even doubtful if it could have been 

 made to appeal to them on the ground of ration- 

 ality. 



The principle of relativity is the direct outcome 

 of the application of the experimental method, 

 and the full force of its appeal is based on our 

 absolute confidence in the metaphysical concept 

 of reality which is the ground and rea.son of that 

 method. The experimental method has taken 

 possession of the modern mind, and it assumes 

 for us something like the unmodifiable character 

 of an instinct. If experiment proves a certain 

 velocity to be constant under conditions which 

 require us to predict its variation ; if experiment 

 shows the movement of a source of light to be 

 without the expected effect on the velocity of pro- 

 pagation — well, it is our concept of the nature of 

 reality which must adapt itself to the experiment. 

 The prediction is based on the concept that space 

 and time provide an absolute system of refer- 

 ence; the null result of the experiment negatives 

 that concept, and henceforth space and time are 

 " shadows " ; they must vary, because under 

 varying conditions velocity is constant. 



Those who affirm that the principle of rela- 

 tivity is purely mathematical, and not meta- 

 physical, and, therefore, resent the intrusion of 

 metaphysics into the discussion of its equations, 

 conceive the principle to be purely methexlological, 

 to be concerned only with abstract quantitative 

 measurement, and merely to substitute a 

 very complex and difficult set of cquation- 

 formulfe for a discarded simpler one, in 

 the interest of greater precision and accuracy 

 alone. Those who take this view seem to me to 

 misapprehend the significance of the principle. It 

 is to Ik- understood only when taken in its his- 

 torical connection with the metaphysical con- 

 structions of the great philosophers. 



Since Descartes, the speculations of philosophy 

 have centred round the concepts of substance and 



