THE QUARTERLY 



JOURNAL OF SCIENCE 



JANUARY, 1871. 



I. ON DOUBLE SPECTRA. 

 By W. Marshall Watts, D.Sc. 



tT is now just ten years since the first principles of spec- 

 trum analysis were enunciated by Bunsen and Kirchhoff. 

 It would, probably, be difficult to find any ten years in 

 the history of science which have yielded such a glorious 

 harvest of results. In these ten years we have not only become 

 acquainted with four elements, constituents of the earth's 

 crust, whose existence was previously not even suspected, 

 but we have been enabled to extend our analysis beyond the 

 earth, to far distant stars, and to learn the composition of 

 our sun and of other suns, whose distance from us is so 

 great that, in comparison with it, the immense space which 

 stretches between the sun and the earth sinks into insig- 

 nificance ; nay, more, we have been able to detect star- 

 motions, which it is beyond the power of the telescope to 

 reveal — have gained extensive and important insight into 

 the physical constitution of our great luminary, the sun, 

 since the spectroscope enables us to observe any day, at 

 leisure, phenomena which, without the spectroscope, can be 

 studied only during the rare occurrence of a total eclipse, 

 and are even able to look back into past time and to trace 

 the history of the system of which our planet forms a part. 

 But we have in these ten years not only learnt new modes 

 in which this mighty weapon of research may be employed, 

 but have also made not less important progress in other 

 directions, viz., in the establishment on sure ground of the 

 principles on which the analysis rests, and in the definition 

 of the precise conditions under which it is applicable. It 

 must not be forgotten that every conclusion to which experi- 

 mental science comes is a result of induction, and the danger 

 of too hasty generalisation is manifest when we come to 

 work with principles so potent as those of spectrum analysis. 

 If, for example, we conclude because the spectrum of solar 

 light contains a dark line of the same refrangibility as the 

 dark line which can be obtained artificially in the spectrum 

 of incandescent sodium vapour, that the sun's atmosphere 



VOL. VIII. (O.S.) — VOL. I. (N.S.) B 



