280 Prof. J. N. Lockyer. [June 21, 



roughly one for each day of the period are given in Plate 1. These 

 have been enlarged about three times from the original negatives.* 



As it is a matter of great difficulty to mount a series of such photo- 

 graphs showing the exact coincidences of the lines, in comparing the 

 different spectra in the plates some allowance must be made for the 

 slight differences in scale. Further, it is right to add that probably 

 some of the fainter lines shown in the photographs are artificially 

 produced by the process of enlargement, but the real lines will be 

 readily identified by their appearance in more than one spectrum; the 

 lines of particular interest are indicated in Plate 2 (p. 283). 



The light curve which forms part of Plate 1 is constructed after 

 Argelander's drawing, f and the dotted lines drawn from the spectra 

 to the period scale indicate the relation of each photograph to the 

 light curve. 



I proceed to state, step by step, the results of the preliminary 

 examination of the photographs, and to indicate the spectral pheno- 

 mena on which they are based. 



1. The spectrum is constant at the same interval from principal 

 'minimum. 



Apart from the slight differences which seem to be accounted for 

 by differences in the atmospheric conditions and consequently in the 

 quality of the negatives, the spectrum 'appears to be the same at the 

 same interval from minimum. The photographs reproduced in 

 Plate 1 have been selected as being specially suitable for reproduc- 

 tion, but at most of the phases duplicates which are practically iden- 

 tical have been obtained. 



2. The kinds of variation shown on the photographs are as follows : 



(a.) Periodical changes in the relative intensities of the lines. 

 (6.) Periodical doublings of some of the dark lines, 

 (c.) Periodical changes in the positions of the bright lines with 

 respect to the dark ones. 



3. There are two bodies involved giving dark line spectra. 



On reference to Plate 1 it will be seen that at, and just before and 

 after the second maximum, some of the dark lines are doubled. This 

 indicates two sources of light giving dark line spectra and moving 

 relatively to each other in the direction of the line of sight. When 

 the relative movement in the line of sight is zero, none of the lines 

 are doubled. The latter condition occurs about the time of the two 

 minima. 



4. The maximum relative velocity of the two dark line components in 

 the line of sight is about 156 miles per second. 



* This plate is not given in the " Proceedings," as it is very difficult to reproduce 

 it on so small a scale. 

 t De Stella )3 Lyrse Disquisitio. 



