Septembee 19, 1913] 



SCIENCE 



391 



sense establishes itself, it seems as if even 

 time would become discontinuous and be 

 supplied in atoms, as money is doled out in 

 pence or centimes instead of continuously 

 — in whicli case our customary existence 

 will turn out to be no more really continu- 

 ous than the events on a kinematograph 

 screen — while that great agent of continu- 

 ity, the ether of space, will be relegated to 

 the museum of historical curiosities. 



In that case differential equations will 

 cease to represent the facts of nature, they 

 will have to be replaced by finite differ- 

 ences, and the most fundamental revolution 

 since Newton will be inaugurated. 



Now in all the debatable matters of 

 which I have indicated possibilities I want 

 to urge a conservative attitude. I accept 

 the new experimental results on which 

 some of these theories — such as the prin- 

 ciple of relativity — are based, and am pro- 

 foundly interested in them, but I do not 

 feel that they are so revolutionary as their 

 propounders think. I see a way to retain 

 the old and yet embrace the new, and I 

 urge moderation in the uprooting and re- 

 moval of landmarks. 



And of these the chief is continuity. I 

 can not imagine the exertion of mechanical 

 force across empty space, no matter how 

 minute; a continuous medium seems to me 

 essential. I can not admit discontinuity 

 in either space or time, nor can I imagine 

 any sort of experiment which would justify 

 such a hypothesis. For surely we must 

 realize that we know nothing experimental 

 of either space or time, we can not modify 

 them in any way. We make experiments 

 on bodies, and only on bodies, using 

 "body" as an exceedingly general term. 



We have no reason to postulate any- 

 thing but continuity for space and time. 

 We cut them up into conventional units for 

 convenience' sake, and those units we can 

 count; but there is really nothing atomic 



or countable about the things themselves. 

 We can count the rotations of the earth, 

 or the revolutions of an electron, or the 

 vibrations of a pendulum, or the waves of 

 light. All these are concrete and tractable 

 physical entities; but space and time are 

 ultimate data, abstractions based on ex- 

 perience. We know them through motion, 

 and through motion only, and motion is 

 essentially continuous. We ought clearly 

 to discriminate between things themselves 

 and our mode of measuring them. Our 

 measures and perceptions may be affected 

 by all manner of incidental and trivial 

 causes, and we may get confused or ham- 

 pered by our own movement; but there 

 need be no such complication in things 

 themselves, any more than a landscape is 

 distorted by looking at it through an irreg- 

 ular window-pane or from a traveling 

 coach. It is an ancient and discarded 

 fable that complications introduced by the 

 motion of an observer are real complica- 

 tions belonging to the outer universe. 



Very well, then, what about the ether, is 

 that in the same predicament? Is that an 

 abstraction, or a mere convention, or is it a 

 concrete physical entity on which we can 

 experiment ? 



Now it has to be freely admitted that it 

 is exceedingly difficult to make experiments 

 on the ether. It does not appeal to sense, 

 and we know no means of getting hold of 

 it. The one thing we know metrical about 

 it is the velocity with which it can trans- 

 mit transverse waves. That is clear and 

 definite, and thereby to my judgment it 

 proves itself a physical agent; not indeed 

 tangible or sensible, but yet concretely real. 



But it does elude our laboratory grasp. 

 If we rapidly move matter through it, 

 hoping to grip it and move it too, we fail : 

 there is no mechanical connection. And 

 even if we experiment on light we fail too. 

 So long as transparent matter is moving 



