November 12, 19 14] 



NATURE 



281 



Those who wish to read a simple and brief 

 account of the main facts about the sun, planets, 

 asteroifls, comets, and stars (normal and abnormal) 

 cannot do better than peruse this book. A general 

 idea of the universe will be easily obtained, and 

 some notion of the vast subject of astronomy will 

 be gained. The illustrations are reproduced from 

 some of the finest photographs eVer taken, and 

 they are well distributed and chosen. A brief 

 bibliography of books more or less popular is 

 given at the end ; also a chronology of the main 

 events of astronomy from very early times to 

 1908. With regard to the latter the author seems 

 to think that such a table is unique or " what the 

 compiler has never happened to see before." To 

 take only one instance, he has evidently never 

 seen Miss Gierke's admirable "History of Astro- 

 nomy during the Nineteenth Century," at the end 

 of which is a chronological table stating the main 

 astronomical advances from 1774 to 1893. 



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.} 



Fizeau's Experiment and the Principle of Relativity. 



Mr. Cunningham seems determined that the prin- 

 ciple of relativity shall assert supremacy as regards 

 the Fizeau effect on the velocity of light, arising from 

 motion of the transmitting medium. This is a first- 

 order effect, while the principle of relativity was 

 invented to account for the second order. Up to the 

 first order the affairs of moving media hang entirely 

 on the ascertained fundamental relations of electro- 

 dynamics ; and any type of relativity which does not 

 agree with the formula thence deduced must, if sub- 

 stantiated, upset our whole system of ideas. 



As regards the arguments adduced: (i) is of 

 course a slip of the pen — he is thinking of steady 

 viscous flow in a pipe, whereas the standard 

 measures for ordinary turbulent flow are in general 

 agreement with the values quoted from the experi- 

 ment; while as regards (2), it seems certain that the 

 experimenters must have had before them the effect 

 of variation of velocity across the section of the pipe, 

 which would curve the fronts of the waves and thus 

 broaden the interference bands, but not much if only 

 the central part of the pipe, in which the velocity 

 varies but slightly, is employed. The shift at the 

 centre of the band-system would depend on the axial 

 velocity only. Joseph L.xrmor. 



Cambridge, November 2. 



There was no intention in the letter to which Sir 

 Joseph Larmor refers (Nature, October 29, p. 226) 

 of suggesting that the principle of relativity and the 

 fundamental relations of electro-dynamics were at 

 variance with respect to this matter. Certainly I 

 have always taken them to be completely at one. 

 But it has been raised in several quarters as an 



NO. 2350, VOL. 94] 



objection to both points of view that the extremely 

 careful experiment of Michelson and Morley was not 

 in complete agreement with theory. They found the 

 convection coefficient for water to be 0-4341002, while 

 theory (taking account of dispersion according to 

 I Lorentz's formula) gives 0-451, which is very near the 

 limit of possible error. 



Sir Joseph Larmor makes it clear that the sug- 

 gestions which I made as to the possible origin of 

 the discrepancy are untenable, and, i understand, 

 feels that it must lie m the inherent difliculties of the 

 experiment. One of these is the determination of 

 I the axial velocity in terms of the mean. The experi- 

 i mental device could not be expected, I think, to give 

 i a result correct to one or two per cent., inasmuch as 

 the pressure gradient actually measured is partly the 

 ! normal pressure gradient of the undisturbed flow, 

 I and partly that due to the disturbance produced by 

 I the insertion of the gauge-tubes. 



But, as a matter of fact, the discrepancy is not so 

 great as has been thought. A recalculation from 

 the figures and formulae of Michelson and Morley, 

 which Sir Joseph Larmor has now checked for me, 

 gives for the experimental value of the coefficient 

 0442, with their estimated possible error of ±0-02.' 

 Thus the discrepancy (0-009) with the value given by 

 Lorentz's formula (including dispersion) proves to be 

 less than half the possible error. 



E. CtNMN(,H\M. 



Wireless Signalling for Shipping in War Time. 



The advantage which wireless telegraphy has been 

 in enabling the whereabouts of vessels at sea to be 

 known from day to day, is now in abeyance, I believe, 

 on account of the war, because of the messages being 

 liable to be read by the enemy. 



But if a system of signalling false latitudes and 

 longitudes bv a code prearranged by the owners, 

 different for each craft, and variable from day to day, 

 easv and simple to translate by reference to the copy 

 kept, then wireless signalling could go on as freely 

 as before, without danger of the code being captured 

 by the capture of any craft. 



To illustrate the method of this, let us suppose a 

 firm of owners have two ships, the Ariadne and Ocean 

 Bird. A list of dates has been given to each captain 

 with instructions to falsify positions on the respective 

 dates (London time) with additions and subtractions 

 as found in the list in opfxjsition to the dates, thus : — 



Ship 



Ncv. I 



Ship 



Nov. I 



The owners of these ships could interpret the wire- 

 less messages received through the exchange, and the 

 Admiralty be quickly notified of a disappearance with 

 approximate latitude and longitude. 



A. BOW.\IAN. 



144 Well .Street, Hackney. 



