October 7, 1898.] 



SCIENCE. 



455 



the computations to each catalogue are sub- 

 stantially those assigned by Professor Lewis 

 Boss in his report on similar work for the 

 Northwest Boundary Survey in the seventies. 



Professor Pickering presented a paper by 

 Mrs. M. Fleming on the ' Classification of 

 the Spectra of Variable Stars of Long 

 Period.' In the classification of the spec- 

 tra of stars in the Draper Catalogue the 

 letter M was assigned to stars having spec- 

 tra of the third type. Later these stars 

 were subdivided into four groups. Ma, Mb, 

 Mc and Md, of which a Orionis, a Herculis, 

 — 2°3653, and o Ceti were given as examples. 

 In the last of these stars the hydrogen lines 

 are bright, and this spectrum has been 

 shown to be characteristic of variable stars 

 of long period ; indeed, by this peculiarity 

 in the spectrum about a hundred new vari- 

 able stars have been found. A further ex- 

 amination of these spectra shows that they 

 can be further subdivided into eleven 

 groups. A classification made from an ex- 

 amination of the continuous spectrum, the 

 comparative brightness of the hydrogen 

 lines being also carefully estimated, always 

 assuming the brightness of H^ as 10, the 

 first class, of which R Lyncis is the typical 

 star, showed a spectrum resembling a Tauri, 

 and having H/? and H^ strongly bright, and 

 nearly equal, while H5 was barely visible. 

 The last group, of which R Leonis is a 

 typical star, shows a continuous spectrum 

 like —2° 3653 with H/3 invisible, H^ barely 

 visible and H5 strongly marked. 



The original photographs were exhibited 

 and examined by the Conference. 



Professor G. W. Myers presented ' A 

 Study of the Light Variation of U Pegasi.' 

 The hypotheses assumed were : (1) The 

 light curve given by Pickering in Harvard 

 College Observatory Circular No. 23 repre- 

 sents the true nature of the light change of 

 this star near enough to determine provis- 

 ional orbital elements, and (2) The light 

 change is capable of explanation on the sat- 



ellite theory. After the discussion showed 

 the desirability of introducing a flattening, 

 the further assumption was made : (3) The 

 two components are deformed by tidal in- 

 fluence into similar ellipsoids of revolution. 



Equations were then derived connecting 

 the light curve with the ' elements ' defining 

 the system whereby the latter might be de- 

 rived from the former, and vice versa. From 

 these equations the following results were 

 obtained : The light curve of TJ Pegasi, 

 given in circular No. 23, is satisfactorily 

 represented by the satellite theory with a 

 circular orbit, whose radius differs but 

 slightly from the sum of the major axes of 

 the components, and whose inclination is 

 very nearly 90 degrees. The smaller com- 

 ponent has a brightness of about 0.77, and 

 a radius of 0.78 of the larger, while a slight 

 flattening of the disks probably exists. Ac- 

 curacy of present observations does not 

 warrant a least square adjustment of the 

 outstanding residuals. There is some rea- 

 son to believe in the existence in this sys- 

 tem of the apioidal form of Poincar6. 



Professor E. C. Pickering said it was as- 

 tonishing how slight were the differences 

 between observed and theoretical values of 

 the brightness, and this agreement was con- 

 firmatory both of the observations and the 

 theory. Mr. H. M. Parkhurst could not 

 accept the period assumed, and questioned 

 the accuracy of determinations of minima 

 diifering by but .15 of a magnitude. Pro- 

 fessor M. B. Snyder held that the relation 

 of Professor Myers' present work to that of 

 Doctor See on the evolution of binaries was 

 quite remarkable, and, seeing that Doctor 

 See's prediction of variation of eccentricity 

 was just what Doctor Myers had required 

 as an hypothesis in treating the similar 

 variable /3 Lyrse, the results were anything 

 but discouraging to all the hypotheses in- 

 volved in their interesting work. 



Professor George E. Hale, exhibiting a 

 set of remarkable photographs, presented a 



