Makch 6, 1908] 



SCIENCE 



389 



is impossible" (p. 114). "All the scientist 

 creates in a fact is the language in which he 

 enunciates it" (p. 121). " Since the enuncia- 

 tion of our laws may vary with the conven- 

 tions we adopt, since these conventions may 

 modify even the natural relations of these 

 laws, is there in the manifold of these laws 

 something independent of these conventions 

 and which may, so to speak, play the role of 

 the universal invariant? ... In any case a 

 minimum of humanity is necessary" (pp. 

 127-8). "All classification supposes the ac- 

 tive intervention of the classifier" (p. 135). 

 " Sensations are therefore intransmissible, or 

 rather all that is pure quality in them is in- 

 transmissible and forever impenetrable. But 

 it is not the same with relations between these 

 sensations" (p. 136). "Nothing is objective 

 except what is identical for all; now we can 

 only speak of such an identity if comparison 

 is possible, and can be translated into a 

 ' money of exchange ' capable of transmission 

 from one mind to another. Nothing, there- 

 fore, will have objective value except what is 

 transmissible by ' discourse,' that is, intel- 

 ligible " (p. 137). "All that is not thought 

 is pure nothingness; since we can think only 

 thought and all the words we use to speak of 

 things can express only thoughts, to say there 

 is something other than thought is, therefore, 

 an affirmation which can have no meaning " 

 (p. 142). 



It is plain enough, from these representative 

 and characteristic selections, that M. Poincare 

 has not acquired familiarity with psycholog- 

 ical investigation; that, as yet, he has not 

 compelled himself to think through to a defi- 

 nite, coordinated, basis in epistemology ; that 

 his logical methods tend to gloss the secondary 

 character of symbolism; and, above all, that 

 he has not clarified the ultimate metaphysical 

 problem immanent in his acute dialectics. 

 But of these limitations, as the professional 

 philosopher will at once see them, I incline to 

 make light. For it is an immense gain that 

 M. Poincare should have insisted, not merely 

 upon the existence of such riddles, but upon 

 their fundamental import for an evaluation of 

 scientific modes of presentation. 



The book ought to be in the hands of all 

 who desire to " mix their colors with brains." 

 R. M. Wenley 

 University of Michigan 



Atlas of Ahsorption Spectra. By H. S. 

 Uhler and E. W. Wood. Carnegie Insti- 

 tution of Washington, Washington, D. C, 

 1907. 



" To furnish graphical representations, on 

 a normal scale of wave-leng-tlis, of the ab- 

 sorption spectra, both in the visible and ultra- 

 violet regions, of a reasonably large number 

 of compounds," is stated by the authors as 

 their chief object in producing this book, and 

 with the exception of the fact that their 

 spectrograms do not extend into the red, their 

 object has been very well attained. 



The book opens with a two-page introduc- 

 tion by Professor Wood, which is followed by 

 eighteen pages including descriptions of the 

 apparatus used, spectrograph, sources of light, 

 photographic materials, explanation of the 

 tables, etc. The tables occupy about forty 

 pages, and give, in systematic form, the re- 

 sults obtained for 147 aniline dyes and some 

 of their related organic compounds, and 36 

 miscellaneous absorbing media, chiefly inor- 

 ganic salts. Twenty-six plates, 102 figures, 

 positives of the spectra observed, complete the 

 book. 



The dispersing apparatus used was a con- 

 cave grating of 98.3 cm. radius, the ruled 

 surface of which was 1.96 cm. by 5.36 cm. 

 Most of the photographs were taken on cellu- 

 loid films, sensitized with Seed's " L-ortho " 

 emulsion. A few photographs were taken on 

 Cramer's Trichromatic plates, for the orange 

 and red regions up to about .63 ix. Most of 

 the plates extend from about .20 ju, or .22 /a 

 to about .59 /jL or .60 /x, where the Seed plates 

 cease to be sensitive for normal exposures. 



A Nernst glower carrying .8 ampere on a 

 104 volt 133 cycle circuit, furnished a con- 

 tinuous spectrum down to about .32 fj, or .34 [i. 

 A spark between electrodes, one of sheet brass 

 and the other of equal parts of zinc and 

 cadmium, furnished a bright line spectrum 

 from about .2 /x up. The spark spectrum was 

 cut off from the plate by a movable screen 



