200 



NATURE 



[May 4, 1916 



be a minimum at minimum deviation, for if we turn 

 through minimum deviation the spectrum is 

 shortest there." But on the ordinary definition 

 of the dispersion this is far from true. Perhaps 

 the author's definition of dispersion is to be pre- 

 ferred. The author alleges, on p. 252, that it 

 is difficult to show in the laboratory the reversal 

 of the sodium lines. If he will adopt the follow- 

 ing plan he will, on the contrary, find it very 

 easy, even as a lecture demonstration. Use a 

 hand-feed arc lamp. Let the lower carbon be 

 hollowed out so as to form a sort of small 

 crucible; but let a slight V-notch be dut in its 

 rim on the side towards the projecting lens- 

 system of the lantern. Let the upper carbon be 

 thin and pointed and set to strike the arc by con- 

 tact with the rim on the opposite side. Put a 

 pellet of sodium in the "crucible," and then move 

 the top carbon down and up several times so as 

 to strike the arc repeatedly. A continuous 

 spectrum is evoked, accompanied usually by 

 bright lines, including the D-line; but the D-line 

 at once changes to a black line, since the light 

 has to pass through the mass of sodium vapour 

 which IS slowly pouring over the V-notch. 



In the chapter on the later spectroscopy Row- 

 land's photographic charts, Balmer's series, the 

 work of Kayser and Runge, and that of Stark, 

 Zeeman, and Michelson, are admirably described 

 and summarised. The chapter on infra-red and 

 X-rays is also admirable, but the early work of 

 Crookes, which led up to the radiometer, is ig- 

 nored. Chapter xx., on lamps and illumination, is 

 less satisfying. Surely the estimate of 200,000 

 candles per square inch for the intrinsic brilliancy 

 of the crater of an arc lamp is too high. 



The fourth part, the mathematical theory of 

 light, is a very able and very welcome feature of 

 the work, though it is not all easy reading. It 

 deals with the propagation of single pulses and 

 groups of waves ; the modern notion of the true 

 function of prisms, not as sorters-out of hypo- 

 thetically pre-existing trains of periodic waves, 

 but as the manufacturers of these trains out of 

 miscellaneous and utterly irregular impulses ; the 

 electromagnetic theory of light; the experiments 

 of Hertz ; the problems of reflexion and refraction ; 

 the theory of dispersion ; the theory of radiation ; 

 and the pressure of light. A pregnant chapter 

 on the relative motion of matter and ether, in 

 which the celebrated paradoxical experiment of 

 Michelson and Morley forms the pivot of the argu- 

 ment, brings the book to the close with the remark 

 that the Michelson-Morley experiment is a some- 

 what narrow basis on which to rear such a struc- 

 ture as the "relativity" doctrine of Einstein. It 

 is indeed. 



To many of the chapters Dr. Houstoun has 

 appended series of questions and problems. 

 These are excellent, being real problems of optics, 

 and not, as in the majority of college text-books, 

 mere mathematical puzzles. There is a reality 

 and freshness about them that is wholly com- 

 mendable. 



The tables included at the end of the book are 

 all too short. But they are satisfactory com- 



_ - _ _ _ ^T 



pared with the index. One looks in vain for manj 

 things. The index contains no reference to aper- 

 ture. Angstrom's unit, crossed prisms, diffuse 

 reflexion, index of refraction, luminosity, lumines- 

 cence, persistence of vision, power of a lens, re- 

 fraction, or selective radiation; and the inquirei 

 who wants to know the significance of jx or of /xji 

 will vainly hunt for the footnote on p. 243 01 

 that on p. 298, where these mysteries are revealed, 

 In an important text-book such as this an indes 

 ought not to be left to a compiler who does nol 

 grasp what are the good things that must not b( 

 left uriindexed. One misses even any reference 

 to some of the best and most instructive thing* 

 in the lx>ok, the original researches of the authoi 

 himself, which are to be found on pp. 299, 324 

 and 350. The publishers ought at once to scraf 

 the index without waiting for the second edition 

 which is certain to be called for at no distani 

 time. S. P. T. 



OVR BOOKSHELF. 



The Moon: Considered as a Planet, a World, 

 and a Satellite. By J. Nasmyth and J. Car- 

 penter. Cheap edition. Pp. xix+315. (London: 

 J. Murray, 1916.) Price 2s. 6d. net. 

 It is a pleasure to direct attention to the issue, al 

 an extraordinarily low price, of a complete 

 edition of Nasmyth and Carpenter's classical wxjrl- 

 on lunar physiography. Accustomed as we are o 

 late years to cheap editions, this reprint appearij 

 to us to present really exceptional value. Th»i 

 work first appeared forty-two years ago, and w. 

 reviewed in Nature of March 12, 1874. That t; 

 appreciative tenor of that review was. entire 

 deserved is sufficiently evidenced by the fact th: 

 four editions have been issued. Nevertheless, > 

 may not be out of place to quote a fairly recen 

 French endorsement : — 



**Au point de vue pittoresque, aucune repr 

 sentation precedente ne pouvait donner une me 

 leure id6e de ce que Ton voit au telescope que 1* 

 reliefs de Nasmyth. Les photographies actuelli 

 sont plus exactes, mars elles sont loin d'atteindi 

 le charme des planches de cet ouvrage qu'on i 

 se lasse point d'examiner." 



It so happens that the review copy of the fir 

 edition has somewhat often been in the hands ' 

 the present writer, and as the illustrations a 

 such an important feature, it is satisfactory to ' 

 able to state that the reproductions in this late 

 edition compare favourably with the original: 

 The text, written when distinction of literary st} 

 could be found even in books of science, can st, 

 be read with profit and with pleasure. 



H. E. G. 



Graphics and Structural Design. By Prof. H. -. 



Hess. Second Edition. Pp. viii + 435. (N< 



York : J. Wiley and Sons, Inc. ; Londoi 



Chapman and Hall, Ltd., 1915.) Price 12s. t 



net. 



The first edition of this book appeared in ig'^ 



The author was formerly designer and compu" 



for the Pencoyd Iron Works and the AmeriC 



