102 



Mr. J. Norman Lockyer. 



Tims, although in regard to the general character of the light 

 changes, /3 Lyrse must be included in this class, it stands apart from 

 the other members which have so far been spectroscopically ob- 

 served in having a spectrum which includes a number of bright 

 lines ; a detailed study of the spectrum has also shown* that the 

 lines are periodically doubled, and that part of the variability is due 

 to eclipsing, although the minima are not sharply marked like those 

 of Algol. 



Again, in the great majority of the variables of Class IY there is, 

 as a rule, a rapid rise to maximum, followed by a slower descent to 

 minimum, as in the variables of Class II. 



Excluding /? Lyrae, which, as I shall show, is spectroscopically 

 very different from the others so far examined, it will be convenient 

 to refer to the remaining variables of Pickering's Class IV as those 

 of the S Cephei class, and it is with some of these that the present 

 paper is concerned. 



The Kensington Vhotographs. 



On attempting to continue the investigations of variability to 

 which reference has been made, I found that the available spectro- 

 scopic data with regard to the t Cephei class were very meagre, and I 

 therefore determined to investigate the spectra photographically, so 

 far as the means at my disposal would permit. The work was 

 accordingly commenced at Kensington in the early part of 1894, and 

 as some interesting facts have been already brought to light, I have 

 decided, although the research is not yet completed, to communicate 

 them to the Society, more especially as my instrumental means only 

 permit the study of a small number of the stars. 



The instruments employed in the inquiry were those named A and 

 B in a previous communication on the " Photographic Spectra of some 

 of the Brighter Stars ;"f the former consists of a prism of 1\ °, placed 

 in front of a 6-inch photographic object-glass, and the latter of a 

 prism of 45° used in conjunction with the same telescope. 



The spectra of most of the variables of this class are too faint to 

 be photographed with the great dispersion of B, so that, in some 

 cases, A has of necessity been solely employed. Brief particulars as 

 to the variability of the five stars, the spectra of which have been 

 already photographed, and a statement of the number of photographs 

 taken in each case, are given below. 



* 1 Roy. Soc. Proc.,' vol. 56, p. 280, 1894. 

 + ' Pbil. Trans.,' 1893, vol. 184, A, p. 678. 



