September 22, 1904] 



NA TURE 



503 



cocoa-nut oil in butU-r. No reference, however, is 

 made to the Miint/ and Coudon method of estimating 

 the latter adulterant. 



These are but a few out of many points of interest 

 which one notes on looking through the book. A 

 number of new illustrations appear, including some, 

 which might be improved upon, of lard, cholesterol, 

 and phytosterol crystals. There are plenty of refer- 

 ences to original sources, and the information generally 

 is brought well up to date, several papers issued in the 

 present year being laid under contribution. 



" .\duIteration," says Dr. Lewkowitsch, " has 

 almost become a fine art." No doubt it has; and in 

 the silent, ceaseless struggle between the cunning of 

 the adulterator and the skill of the analyst such works 

 as the- present play an important part. They are very 

 helpful to the former individual, certainly. But to the 

 latter they are invaluable. C. Simmonds. 



STOKES'S MATHEMATICAL AND PHYSICAL 



PAPERS. 

 Mathematical and Physical Papers. By the late Sir 

 G. G. Stokes. Vol. iv. Pp. viii + 378. (Cam- 

 bridge : The University Press.) Price 15.?. 

 IT was on all grounds fitting that the continuation 

 of this reprint should be entrusted to Prof. 

 Larmor. The energy with which he has addressed 

 himself to the work is shown by the fact that, although 

 it is little more than a year since the death of his great 

 predecessor, we already have a new volume in our 

 hands, containing, in addition to the text, some 

 valuable annotations and a selection from some very 

 interesting correspondence. 



The papers here reproduced range in date from 1853 

 to 1876; they are about forty in number, and, as a 

 rule, are shorter and more restricted to special points 

 than is the case in the previous volumes. There are, 

 however, some notable exceptions. From the mathe- 

 matical point of view the most considerable is the 

 memoir " On the Communication of Vibration from a 

 Vibrating Body to a Surrounding Gas." Perhaps the 

 highest testimony to the excellence of this investigation 

 is that Lord Rayleigh, who usually transforms and 

 illuminates what he touches, in this case found, as 

 he tells us, no better course open to him than to print 

 page after page verbatim in his "Theory of Sound." 

 The memoir is important, historically, not solely for 

 the interest of the particular phenomenon which it 

 explains, but as leading the way for a whole series 

 of investigations in acoustics, optics, and electricity, 

 in which we have to deal with waves diverging from 

 point- or line-sources. Especially characteristic of the 

 author is the labour expended with a view of reducing 

 the results to a definite numerical form. From another 

 point of view the paper may be regarded as forming 

 one of the long series (some other members of which 

 fall in the present volume) in which Stokes attacked 

 the difficulties of the Bessel functions ; other methods 

 of dealing with these have since been devised, but it 

 is mainly through his labours that these functions have 

 become real and intelligible instruments of the mathe- 

 NO. I 82 I, VOL. 70] 



matical physicist, instead of merely abstract analytical 

 expressions. 



We also find in this volume the classical " Report 

 on Double Refraction," presented to the British 

 Association in 1862. This has entered into so many 

 discussions, that it is unnecessary to refer to it in detail. 

 Although elastic theories of light no longer excite the 

 same interest, the report is still worthy of careful study, 

 not only on intrinsic grounds, but also as a masterpiece 

 of criticism, and as an embodiment of the clear and 

 judicial mind of its author. 



Among experimental investigations, we may note 

 the very important paper " On the Long Spectrum of 

 the Electric Light," and the verification of Huyghens's 

 law of refraction in uniaxal crystals, which has served 

 as a touchstone of optical theories. 



A short, but extremely acute, paper " On the Effect 

 of Wind on the Intensity of Sound," read before the 

 British Association in 1857, was unfortunately un- 

 noticed and forgotten until the explanation was re- 

 discovered, and extended so as to include the effect of 

 variations of temperature, by Osborne Reynolds in 

 1874. 



It will not be supposed that the numerous other brief 

 memoirs which we are obliged to pass over without 

 special mention are unimportant. To the scientific 

 worker the value of such a collection often resides 

 chiefly in these minor investigations, which are other- 

 wise in danger of being overlooked, as in the instance 

 just referred to. 



As has been already mentioned, the editor has 

 appended a few notes, chiefly of a historical character. 

 This delicate task has been exercised with great judg- 

 ment and restraint. He has also included a most 

 interesting correspondence between Stokes and Thom- 

 son on the early history of spectrum analysis. It is- 

 clear that long before Kirchhoff's first publication on 

 the question Thomson was in possession of the leading 

 ideas of the subject, and foresaw its wonderful possi- 

 bilities, and that he had, moreover, publicly expounded 

 these things in his lectures at Glasgow. But whilst 

 he is emphatic that he derived his knowledge from. 

 Stokes, the latter is equally positive that his share in 

 the matter was limited to suggestions which he had 

 himself not been able to follow out with the same 

 confidence. The whole correspondence is a lesson of 

 magnanimity on both sides; we feel, as Lord Rayleigh 

 recently expressed it, that the theory of spectrum 

 analysis is practically there, but it would be contrary 

 to the whole spirit of the friendly debate to attempt 

 to analyse further how much of the merit of this pre- 

 vision belongs to one rather than to the other. One 

 point, however, remains indisputably associated with 

 the name of Stokes, viz. the hypothesis that special 

 absorption of light is due to coincidence, or approxi- 

 mate coincidence, of the period of the light waves with 

 a proper period of a molecule. Hypotheses of this 

 kind have played a great part in recent theories of 

 anomalous dispersion and the like ; but there can be 

 no question as to their original source. 



The remaining papers are to be included in a fifth 

 volume, together with the biography by Lord Rayleigh 

 recently issued bv the Roval Society. We are also 



