August 31, 1905] 



NA TURE 



^27 



ten tenth-metres. The region of spectrum studied 

 extends from about x 2500 to x 7000. 



In the production of the arc spectrum, rods of the 

 metal were used as poles whenever possible, though 

 in many cases carbon electrodes were employed, and 

 scraps of metal or salts of the metal volatilised on 

 them. The selection of carbon as electrodes seems to 

 us a very unfortunate one, as it is next to impossible 

 to disentangle the real spectrum of a substance from 

 the structure of the carbon bands. Surely a better 

 method would be to use poles of some inexpensive 

 metal the spectrum of which is a fairly simple and 

 characteristic one, such as zinc, aluminium, or silver. 



Among the spectra represented in the charts are 

 several, such as boron, arsenic, the rare earths, the 

 platinum group, phosphorus, selenium, which are 

 reproduced here for the first time. The previously 

 existing records relating to some of these were very 

 meagre, and those now published will be distinctly 

 useful. For some of the gaseous elements vacuum- 

 tube spectra have been obtained. 



The authors have not given — and it seems un- 

 necessary to do so — complete lists of wave-lengths, 

 but have confined themselves to a selection of the most 

 characteristic lines for each element. The wave- 

 lengths of these are given only to the nearest 

 Angstrom unit or tenth-metre, which is scarcely of 

 sufficient precision for modern spectroscopic research. 

 A chapter of notes is given at the end of the text, 

 touching on such points as varying numbers of lines, 

 kinds of spectra, character of lines, division into 

 pairs, triplets, and series, lines specially prominent in 

 any particular light source, &c. 



Notable amongst the few elements not investigated 

 by these observers is scandium. This is unique among 

 the rarer metals in the prominence of its lines in 

 various celestial spectra — notably the chromosphere 

 and stellar types of intermediate temperature — and a 

 reproduction of its complete spectrum would there- 

 fore have been of interest. 



The reproductions are generally excellent; excep- 

 tion must be taken, however, to that of the solar 

 spectruin, which, apparently included as a reference 

 spectrum, is practically useless. Upon the whole, the 

 production of the atlas is very creditable to the 

 authors, and without being in some ways of so 

 elaborate a nature as Crew's or the recently pub- 

 lished atlas of Eder and Valenta, it will, through its 

 uniform treatment of all the elements investigated, be 

 useful, as the authors surmise, to the physicist, 

 chemist, and astronomer. F. E. B. 



OVR BOOK SHELF. 

 Pn'cis d'Hydraiilique — La Houille Blanche. By 



Raymond Busquet. Pp. viii + 375. (Paris: J. B. 



Bailliere et Fils, 1905.) Price 5 francs. 

 This book forms one of a series of little volumes which 

 are being issued under the title of " Encyclop^die In- 

 dustrielle," and treats of the principles of hydraulics 

 and their applications, which possess an enhanced im- 

 portance in view of the recent great extension of the 

 employment of water-power for industrial purposes, 

 resulting from the discovery that it can be economic- 



NO. 1870, VOL. 72] 



ally transmitted to a distance when converted into an 

 electric current. Thus, by the development of elec- 

 trical transmission, it is now practicable to use water- 

 falls and water stored up in reservoirs, in remote 

 mountain valleys, as sources of power for towns, of 

 which the Falls of Niagara, supplying electrical energy 

 to Buffalo, furnish so notable an instance; and the 

 author has given the name of " La Houille Blanche," 

 or white coal, to this source of power. 



The subject is discussed in five chapters, dealing 

 successively with fundamental laws, flow of liquids 

 in pipes, flow erf liquids in open channels, hydraulic 

 motors, and creation of a fall of water; and the_ text 

 is illustrated by forty-nine diagrams and drawings. 

 The hydraulic problems relating to the utilisation of 

 water-power are solved by aid of arithmetic and simple 

 geometry; and the author's aim has been, by making 

 the book neither purely descriptive nor wholly di- 

 dactic, to render it serviceable to a large number of 

 persons. In the chapter on motors, the different forms 

 of waterwheels and the various types of turbines are 

 described; and, finally, the principle of the hydraulic 

 ram is explained, as being distinct from motors, and 

 yet transforming the fall of water into useful work 

 by raising someof the water to a considerable height. 

 Though reservoirs have been, and are being, formed 

 by constructing high masonry dams across narrow 

 gorges in the valleys of mountain streams, with the 

 object of furnishing water-power, the final chapter of 

 this book relates exclusively to the erection of a 

 masonry weir across rivers, with the necessary sluice- 

 way, closed by wooden panels, for the discharge of 

 floods, by which the ordinary water-level of the_ river 

 is raised so as to enable water to be drawn off into a 

 branch canal for supplying water-power ; and it deals 

 mainly with the requisite calculations of the flow of 

 the river, the discharge through the sluices, the 

 pressure on the panels, the fall available, and the 

 section of the branch canal and of its side retaining 

 walls. The author entertains great expectations as to 

 the future of water-power, and considers that, whereas 

 last century was the century of steam, the twentieth 

 century will be called the age of water-power, or white 

 coal. 



Catalogus Mammaliitni, tarn viventium qiiain fos- 

 silium. By E. L. Trouessart. Suppl. part iv. , 

 Cetacea to Monotremata. (Berlin : Friedliinder 

 and Son, 1905.) Price 8s. 

 We have much pleasure in congratulating the author 

 on the completion of the first quinquennial supple- 

 ment, whereby an absolutely invaluable work is 

 brought well up to date, or, rather, as nearly up to 

 date as is possible in undertakings of this nature. 

 We notice that in the part before us references are 

 given in the case of well-known species to passages in 

 which they have been recently mentioned — a plan 

 which cannot fail to be of the greatest assistance to 

 students. 



In accordance with the recent changes in nomen- 

 clature, the titles adopted for several genera differ 

 from those employed in the original issue, as, for in- 

 stance, Orcinus in place of Orca, on account of the 

 preoccupation of the latter term. In the case of the 

 Edentata, the list of names proposed by Dr. 

 .\meghino for South American Tertiary forms looms 

 very large, and, we fear, occupies much more space 

 than it is really entitled to claim. In this connec- 

 tion it may be noted that the author follows Dr. 

 Wortman in classing the North American Eocene 

 ganodonts as ancestral types of the true edentates, 

 Prof. W. D. Scott's recent opposition to this view 

 probably not having been published in time to receive 



