88^ 



KNOWLEDGE 



[April 1, 1890. 



Thia discovery was closely followed by several similar 

 ones. 1\I. Cornu, early in 1HS4,'- announced with regard 

 to the three great bands of the solar spectrum, A, B, and 

 I>, that the telluric lines of the last named " form two 

 unequal series of double lines, whose channelled appearance 

 immediately recalls that of the telluric groups A and !> ; " 

 that the three bands may be looked upon as forming three 

 harmonic groups analogous to the triplets of magnesium 

 and zinc ; and that the reciprocals of the wavelengths of 

 the homologous lines of the three groups are very nearly 

 in arithmetical progression. A year or so later Prof. 

 Rowland commenced his study of the same bands, and not 

 only confirmed Cornu's observation of their resemblance, 

 but also found that the lines of each band showed an 

 approximation to the same type of structure which Prof. 

 A. Herschcl had remarked in the green carbonic oxide 

 band. But as the lines did not form a series rigidly 



consider the reciprocals — that is, the number of vibrations. 

 Each spectrum is seen to be made up of several series i 

 lines, each of which may be represented with great accuracy 

 by the formula 



^ — A _ _ 



K~ ~ n" n* 

 where X is the wave-length ; A, B, and C are three con- 

 stants ; and n stands for the aeries of numbers from « = 3 

 upwards." ■ 



It will at once be seen that Balmer's formula is only 

 a special case of the above in which C = 0. And it will 



also be seen that as M increases, so , tends to approach A, 



which, accordingly, gives the value of the asymptote of the 

 series. The constant B is ncnrli/ the same for all spectra, 

 whilst the constant C gives the rate of convergence. 



The spectra of the alkalies, however, exhibited other 













harmonic within the limits of accuracy of his measure- 

 ments, he did not publish his results. I 



There is no occasion to review all the attempts which 

 have been made to compel various spectra to give up the 

 secret of their structure. The story of the first complete 

 success in this field has already been told in the pages of 

 Knowledge by the late Editor. (" On the Rhymical Group 

 of Hydrogen Lines visible in many Stellar Spectra,'' by 

 A. C. Eanyard, Kno^\xedge, September, 1891.) Prof. 

 Ames' verification of Balmer's law was followed very 

 quickly by a most important announcement by Profs. H. 

 Kayser and C. Runge with regard to the spectra of the 

 alkalies. These various apectra were all formed, they 

 reported, " in an entirely analogous manner, which is 

 especially manifest if, instead of the wave-lengths, we 



* Comptes Eendtis, Vol. XCVIII., No. 4. 

 t Ames, Phil. Mag., 1890, II., p. 41. 



and more complicated relations than the spectrum of 

 hydrogen. In each case three distinct series were noted. 

 The principal series — the one giving the strongest lines, 

 and those most easily reversed — is the one of longest 

 stride and greatest extent ; and for all the spectra but that 

 of lithium it consists of doublets, the members of which 

 continuaUv approach each other as they approach the 

 red end of the spectrum, the difference of the " wave- 

 number" of the components of any pair being inversely 

 proportional to n. 



But besides the principal series of lines each spectrum 

 contains other subordinate series. Lithium contains two. 

 For both these the constant A as well as the constant B 

 are nearly the same ; in other words, they both approach 



* Sit:, d. Berliner Akai., June 5th, 1890. Profs. Kayser autl 

 Eunge announced in 1888 that they had discovered the general law 

 of spectra, but do not seem to have published it until 1890. 



