ASTKOiNOMY. 141 



tlie second and ninth magnitudes. An important result of his obser- 

 vations is the intimate connection showu between the length of period 

 and the depth of color ot the star ; the very short-period variables are 

 nearly white; those of longer period somewhat redder, the tint grow- 

 ing deeper, the longer the period. 



Several new variable stars have been detected by Chandler, Sawyer, 

 Espiu, and others, and among them are two of more than ordinary in- 

 terest, as they apparently belong to the well-known "Algol-type." The 

 tirst was discovered by Mr. Chandler in the constellation Cygnus (Y 

 Cygni), right ascension =20^^ 47'".5; declination =+340 14' (DM, + 

 340, 4181). Its light varies from 7.1 magnitude to about 7.9 magni- 

 tude, and the period is probably l'^ 11^ 5G'" 48^. The second star re- 

 ferred to, was discovered by Mr. Sawyer in March, 1887, in the constel- 

 lation Canis Major, and as it is the first undoubted variable found iu 

 that constellation, it will probably be known as H Canis Majoris. Its 

 position for 1887 is right ascension —.7'' 14'".4; declination = — IGo 11'. 

 The minimum observed by Mr. Sawyer was 6.7 magnitude, and the 

 j)eriod is 1*^ o'' 15'" 55^. Mr. Chandler has collected the observations of 

 U Oi)hiuchi, of all variables the one with shortest period and the most 

 rapid fluctuations of light, and he finds a curious but well-marked re- 

 tardation iu the increase of brilliancy some half-hour or so after mini- 

 iiuim is i)assed. A similar irregularity has been noticed in the light- 

 curv'^e of S Cancri and occasionally iu that of Algol. 



Mr, Chandler strongly urges the possessors of large refractors to 

 devote a portion of their time to tlie observation of the minima of vari- 

 ables that become too faint for ordinary telescopes, our knowledge of 

 such variables being extremely deficient. Argelander's method of ob- 

 servation is recommended. 



Professor Pickering has in jireparation an index to observations of 

 variable stars which will give for each star the number of observations 

 each year since the discovery of variability. This index will be pub- 

 lished iu volume 18 of the Annals of the Harvard Observatory. 



In an interesting article published iu the Observatory (April, 1888), 

 Miss Gierke has collected a series of notes upon variable double stars. 

 These stars show for the most part a spectrum analogous to that of 

 Sirius, that is of Class I, although single stars of that class hardly ever 

 show any fluctuations iu brightness. Algol-variables, if the eclipse 

 theory of their changes be admitted, make uo exceptions to this rule; 

 Gore's catalogue contains three examples: (5 Oriouis, S Monocerotis, 

 and Y Virginis, and among the well-known doubles are y Virginis, 

 C Bootis, TT Bootis, 38 Gemiuorum, a Piseium, H Serpentis, and /i Scot- 

 pii, and perhaps S Cygni. With few and doubtful exceptions, revolv- 

 ing double stars vary in concert, if they vary at all. The changes of 

 y Virginis illustrate the mode of procedure iu this respect of couples 

 iiitiinsically e(iua1. They alternate in each component, and can thus 

 be detected only by close attention. Each mny bo described as nor- 



