54 Notices respecting New Books. 



increase of the refracting power renders manifest. The method is 

 obviously comparable to the determination of a star by its right 

 ascension and declination. Not only, however, is the zero of Kirch- 

 hoff's scale placed quite arbitrarily, but he also states that " a rela- 

 tion between the numbers on the scale corresponding to the indivi- 

 dual lines and the refractive indices of my prisms does not exist, 

 because the prisms were sometimes placed more exactly than at 

 other times at the angle of minimum deviation for the particular 

 rays." The only natural mode of determining a line, as opposed to 

 an arbitrary mode, is of course the refractive index in some medium 

 of invariable composition, such as water, or bisulphide of carbon. 

 But there are great difficulties in the way of determining refractive 

 indices with sufficient exactness, and, provisionally at least, Kirch- 

 hoff's scale is certain to be adopted. The whole subject would be 

 thrown into great confusion were different scales, each with its own 

 zero, adopted by different observers, like the several systems of lon- 

 gitude ; we earnestly deprecate such a proceeding. 



We must pass rapidly over the second and third parts of the 

 memoir, which treat briefly but pretty completely of " The Spectra 

 of the Chemical Elements," and "The Reversal of the Spectra of 

 Coloured Flames." They contain a review of results which the Pro- 

 fessor had previously published in conjunction with Prof. Bunsen* ; 

 but the purely chemical and mathematical details are omitted and 

 many facts added, especially concerning the spectra of the electric 

 light. The short historical notices of researches in this subject given 

 on pages 6 and 7 are imperfect ; for they do not advert to Wollaston's 

 discovery of the dark lines, and his examination of several kinds of 

 light in 1802, nor to the still earlier observations of Thomas Mel- 

 ville. This latter experimentalist seems to have been the first dis- 

 coverer in the branch of research with which we are occupied : some 

 account of him will be found in the ' Chemical News' for May 3, 1862. 



The solar spectrum, it is well known, is scored by dark lines, 

 while the spectra of the various elements contain bright lines ; and 

 it was not unknown to previous observers (Fraunhofer, Brewster, 

 Foucault, and perhaps others) that the bright yellow sodium line 

 corresponded with Fraunhofer's dark line D. In the course of 

 Kirchhoff and Bunsen's researches, it became manifest that a great 

 number of the solar dark lines correspond to bright lines in the 

 spectra of various metals ; 149 such coincidences are indicated in 

 the Map and Table which we have described. Hence it became very 

 probable that the solar spectrum is the reverse or negative of a com- 

 plicated positive spectrum, in which are combined the spectra of a 

 great many elements. The solar spectrum is related to the spectra 

 of the elements just as a collodiotype negative to its positive copies. 



Reflecting on this remarkable fact, Kirchhoff was led to a further 

 important discovery, developed in his theory of exchanges f. He has 



• See Phil. Mag. S. 4. vol. xx. pp. 1 and 89. 



t Mr. Balfour Stewart, it is well known, had previously arrived at the 

 same important conclusion by a separate path (Trans. Roy. Soc. of Edinb. 



vol. xxii. 1858). See also the paper of M. Angstrom at the commence- 

 ment of this Number of the Phil. Mag. 



