294 Dr. William Hug gins \ • [Feb. 6, 



shown to exist in some acoustic arrangemetius!;^ and which where it 

 exists exalts the intensity of the harir»':^ifc wjose positions nearly 

 fulfil the reri^iiisite condition. I converted tn^lvave lengths into wave 

 frequencies. ... I think it must be accepted that the lines do not lie 

 on, but near a definite curve. This appears to be corroborated by 

 finding that Hi and Gi (hydrogen line near G) are connected har- 

 monially, these rays being exactly the 35th and 32nd harmonies of a 



vibration whose fundamental is (t being the time in which 



light travels a millimetre in air) . . . ." 



Under these circumstances one is led to regard the whole series 

 of lines as due to hydrogen. In this connection it may be stated 

 that Messrs. Dewar and Liveing find that the line of calcium K is 

 more easily reversed than the line at the position of H. 



This spectrum of Vega may be taken conveniently as typical of the 

 whole class of white stars, so that in our consideration of the other 

 stars of this class we shall consider the distinctive features peculiar 

 to each, as modifications, or departures, from this common typical 

 form. To facilitate these comparisons I have distinguished the 

 typical lines by the letters of the Greek alphabet, beginning with the 

 line more refrangible than H. 



In this map (Fig. 4) I have arranged the spectra of five other 

 stars of the white group in their order of change, approximately at 

 least, from the spectrum of Vega. 



I will point out some of the directions in which these changes 

 show themselves, and I will then exhibit upon the screen the photo- 

 graphs themselves of these stars. 



There are principally three directions in which the changes take 

 place : — 



1. In the breadth and greater or less marginal diffuseness of the 

 typical lines. 



2. In the presence or absence of K, and, if present, in its breadth 

 and intensity relatively to H. 



3. In the number and distinctness of other lines in the spectrum. 

 Now in these stars we see modifications in these three directions, 



a successive diminution of breadth of the typical lines, and of the 

 nebulosity at the edges ; the lines become at the same time narrower 

 and defined at the edges. 



In Sirius the lines are about the same thickness as in a LyraB, and 

 the line corresponding to K of about the same fineness. 



In the next star, a Ursae Majoris, we have the same typical group, 

 but the lines are less broad and rather more defined at the edges. 

 There is no fine line at the position of K, but some other lines make 

 their appearance. 



The star next in order is a Virginis. Here the typical lines are 

 still narrower and more defined. K is stronger relatively to H, and 

 numerous lines are visible beyond the last of the typical group. 



In the spectrum of a Cygni the typical lines arc still narrower 



