784 



NATURE 



[August i8, 192 i 



genetic factors, while the cytological study of sex 

 was closely connected with the chromosome-theory 

 of Mendelian heredity. It therefore appears rather 

 tempting to apply the quantitative views of sexual 

 differentiation to the theory of heredity in general. 

 Recently an attempt has been made by the 



writer in "Die quantitativen Grundlagen von 

 Vererbung und Artbildung " (Berlin, J. Springer, 

 1920) to attack the problem of the physiology of 

 heredity from this point of view. A discussion of 

 this would, however, be beyond the scope of this 

 contribution. 



Further Remarks 

 By Sir Oliver 



III. 



Changes of Frequency. 



ACCORDING to the usual presentation of 

 relativity, clocks appear to go slow to a 

 relatively moving observer ; quite irrespective of 

 any Doppler effect, which can readily be allowed 

 for. Their rate would have to be multiplied by 

 the fraction i/)8 or 1^ {i — u^ / c^) ; which means that 

 a clock on the sun seen from the earth, say on 

 December 31 or July i when the motion is exactly 

 transverse, would lose one second in two hundred 

 million, or about sixteen seconds per century. 



But, for testing purposes, we cannot change the 

 motion appreciably, and so we cannot hope to tell 

 if the clock would seem to go quicker if we 

 stopped. Reversal of motion, even if it could be 

 accomplished, would be no good ; the difference to 

 be observed — unlike the Doppler effect — is 

 between motion and rest, or between rapid motion 

 and slow, not between plus and tninus motion. If 

 we had a clock which we could fix at relative rest 

 to ourselves, and yet be sure that it kept time with 

 the one we were observing on some relatively 

 moving body, the comparison might be made. 

 And the revolution or vibration of a radiating 

 atom, (a) on earth, and (b) on sun or star, appears 

 to satisfy the conditions. If source and observer 

 were moving together, there would be compensa- 

 tion ; but if either was moving without the other, 

 there should be an effect, such as by long accumu- 

 lation might be detected. The Mercury effect 

 allowed accumulation for a century or more. The 

 spectrum effect does not allow any accumulation ; 

 whatever can be seen there must be seen instantly, 

 it must depend on what happens in a single period. 

 It is true that a certain train of waves is needed 

 for visibility, and some succession is necessary for 

 interference ; but so short is the series required 

 for interference that position in the spectrum is 

 practically dependent on individual wave-length. 



The value of u^ for the earth's orbit, considered 

 circular, is equal to the sun's gravitational poten- 

 tial at the earth's distance under the inverse 

 square law, say —V; or twice that potential 

 under the direct distance or centrifugal-force law. 

 Hence the slowness to be expected, \/(i— tt^/c^), 

 may be written either i+Y/2c^ or i+V/c^; and 

 the second term is the displacement towards the 

 red which is being looked for. Only, of course, it Is 

 being looked for where the potential is strongest, 

 viz. close to the sun ; for there it is two hundred 

 times stronger than in the neighbourhood of the 



1 Continued from p. 751. 



NO. 2703, VOL. 107] 



on Relativity.^ 



Lodge, F.R.S. 



earth (the radius of the earth's orbit being two 

 hundred times the radius of the sun). It seems, 

 however, that a small fraction of the gravita- 

 tional effect ought to be produced as the result of 

 the earth's motion, even if the sun were nothing 

 but a central source of light. 



The occurrence of the factor 2 is curious, and 

 corresponds with a similar factor in the ray- 

 bending calculation. But I do not now discuss 

 it, because a spectral shift due to transverse motion 

 is doubtful. Space-measuring rods shrink, it is 

 true, but in the direction of motion, not in the 

 sun's direction; so the measured velocity of light 

 from the sun would be constant without any time- 

 correction. Yet it is not easy to see how a clock- 

 discrepancy can be dependent on the direction of 

 motion, apart from the ordinary allowance for 

 light-speed. 



Changes of Inertia and Weight. 



That an electric charge possesses the funda- 

 mental material quality of inertia, by reason of 

 the magnetic field which inevitably is generated 

 when it moves, was first calculated by Sir J. J. 

 Thomson so long ago as 1881. That this electri- 

 cal inertia is a function of speed, so that as the 

 speed of light is approached it ought to undergo 

 a rapid increase of value, was predicted, and its 

 amount reckoned, by both J. J. Thomson and 

 Oliver Heaviside. That the facts of observation 

 were in accord with the prediction, was verified, 

 first by Kaufmann and then by others; while that 

 this subordinate dependence of inertia on speed 

 applies even to neutral atoms of matter, is a con- 

 sequence of the fairly ascertained electrical nature 

 of their constitution. On the theory of relativity 

 the variation of inertia appears to follow, without 

 any electrical theory at all, as a result of changing 

 the frame of reference to moving axes. The addi- 

 tional mass corresponds to the kinetic energy of 

 the moving matter divided by c^. Which sug- 

 gests that the whole mass is probably a demonstra- 

 tion and a result of fine-grained aetherial rotational 

 energy with velocity c. 



It is legitimate, anyhow, to assume as a work- 

 ing hypothesis that the mass of a body is not 

 really constant, but that at the speed u it becomes 

 m = /Swo or m^/ \/(i— u^jc^). 



The speed necessary to display this effect i& 

 usually attained only by electrons and positive 

 nuclei in a vacuum tube, or by aid of spontaneous 

 radio-activity ; but the refinements of astronomy 

 are so great that the planet Mercury is moving 

 fast enough to exhibit some result dependent on 



