"SSO.J on *^ie Ph(>togr(q>hir S2)fctm of ilw Stars. 293 



Now, beyond tlicsa lines '^licrc is auother pair of strong lines in 

 the calcium si)cctruni more relrangiblo. Tlie (inesiiv::;. ^'ises, arc 

 these represented in the star's spectrum by any strong lines"? 'i'liey 

 are not. Two strong lilies arc near them, but do not coincide with 

 them. 



Again, in the ])hotogra2^hic spectrum of hydrogen there is a lino 

 at the position of H. It is seen in Mr. Capron's pliotographic spectra. 

 It is also present in my own i)hotographs. Mr. Lockyer called atten- 

 tion to the coincidence of this line with H in December, 1879. 



In this diagram the photographic spectrum is completed by the 

 addition of the strong lines in the visible one, and laid down according 

 to the scale of wave lengths. Of these tlie first four are certainly 

 lines of hydrogen ; the fifth, H, coincides both with calcium and 

 hydrogen. I suspect in some of my photographs of hydrogen fine 

 lines at the positions of several of the more refrangible lines of the 

 stellar group.* 



That all the lines of this remarkable group are members of a 

 common physical system becomes very highly j^robable indeed, if we 

 convert the wave lengths into their reciprocals, the wave frequencies, 

 and then plot them down, as is done in this diagram (Fig. 5). It then 

 becomes evident tliat they lie on, or very near, a definite curve, a 

 state of things which we could not supj)Ose to happen by chance. 

 Mr. Johnstone Stoney, in a letter to me, remarks on this jioint : — 



" The question whether they lie actually or only near a definite 

 curve is, if I mistake not, of great significance in the theory. If they 

 lie on a curve, obeying any exact mathematical law, their connection 

 must, I think, be attributed to their corresponding to the consecutive 

 overtures of some vibrating system. If, on the other hand, they lie 

 near but not on the curve, this circumstance would sujiport the 

 hypothesis (which seems to accord with other facts) that the visible 

 lines are members of a long series of harmonics, most of the members 

 of which are invisible, those which are seen being those whose posi- 

 tions chance nearly to fulfil a definite state of things which I have 



* Since this discourse was given, Dr. H. W. Vogel has called my attention to 

 a paper on the " Spectrum of Hydrogen ' in the ' Monatsbericlit der Konigl. 

 Acad, der Wissenschaften zu Berlin,' July 10, 1879. In this paper Vogel 

 pointed out coincidence of a line of hydrogen with H ; and among the lines 

 given by him are three others which agree with stellar lines. Vogel's wave 

 lengths for these lines are — 



30G8 H 3834 



3887 3795 



