and the Line-spectra of Hydrogen and Oxygen. 359 



primary substances a and b may now easily be calculated by 

 means of (I.) from the corresponding groups (a) and (b) by 

 multiplication with the factor §, in which operation, unfortu- 

 nately, the errors of the wave-lengths increase in the same 

 proportion. The star-spectra described by Mr. Huggins in 

 the Philosophical Transactions (1880) vol. clxxi.partii. pp. 669- 

 690, those namely of a Aquilee and Arcturus, of which the 

 most refrangible rays are hydrogen-rays, and by comparison 

 with the water-spectrum in accordance with (2) and with the 

 solar spectrum may be confirmed as hydrogen-lines and cor- 

 rected, enable us to fully complete the hydrogen-spectrum 

 as far as known in the ultra-violet portion, and to construct a 

 very complete spectrum of the primary element b. I give 

 here the spectra which I have calculated for the elements (a) 

 and (b) of hydrogen, comparing them with the corresponding 

 Fraunhofer- and chromosphere-lines. 



The wave-lengths of the Fraunhofer-lines were taken at 



once from a good copy of Angstrom's Atlas, and those of the 

 chromospheric lines from Young's second list. I propose to 

 go over the calculations and comparisons as soon as I have at 

 my disposal Thollon's Nouveau Dessin du Spectre Solaire, 

 Piazzi Smyth's ' The Visual Solar Spectrum in 1884/ and 

 especially A. Rowland's ' Photographic Map of the Solar 

 Spectrum.' 



Spectrum of 



Solar spectrum 



element a. 



of Angstrom. 



= 9842-4 





7290-2* 



7290 



6510-2 



6511 



6150-6 



6150-5 



5653-5 &c. 



5653-3 



(With possible errors amounting to a unit of Angstrom's 

 scale. The first two wave-lengths are much more exact than 

 the following ones, because they are derived from the wave- 

 lengths of H a and Hp, as measured with the greatest care by 

 Mendenhall.) 



* Coincides nearly with a telluric line 72903 (Fievez). 



