ASTROPHYSICS IN THE UNITED STATES— BOSLER. 359 



trivances, because of the lack of a convenient standard for the indica- 

 tion of the wave-lengths, it is difficult to use it for the determinations 

 of radial velocities. It is thought that the use of absorbing screens 

 ma}^ do away with this objection last mentioned, now that the sub- 

 stances have been found giving, at ordinary temperatures, fine ab- 

 sorption lines (and not more or less hazy bands). The plates we saw 

 were striking in this respect. This method necessitates, however, a 

 longer exposure. 



Great progress has been made lately in the researches on variable 

 stars. Independent of the direct photometric observations which 

 give comparatively few results, several methods have been devised 

 lor the discovery of new variables. By means of one of the great 

 portrait objectives of 16-inch (40 centimeters) aperture, which Pick- 

 ering has had constructed, eight or ten exposures are made in a 

 series on the same plate at intervals of a half hour; those stars are 

 then easily noted, if such exist, which have varied during the four 

 or five hours' exposure. Variables of short period of the Algol 

 type are thus discovered. Another more general procedure in use 

 at Harvard is to take at two different dates two negatives of the 

 same region of the sky ; a positive is printed by contact from one of 

 them, let us suppose the second, and superposed upon the original 

 first negative. The stars which have not varied are represented — one 

 of the plates being a little more dense than the other — by black dots 

 surrounded by whitish aureoles (or inversely). The images of stars 

 whose light has varied present a different aspect: if their brightness 

 has increased, they appear surrounded by an aureole relatively 

 brighter than that of their neighbors; if the contrary is the case, 

 the aureole is less marked or perhaps wholly absent. By repeating 

 this several times for each region all the variables existing in these 

 regions may be told almost at a glance, even among the immense 

 numbers of stars not undergoing fluctuations. The value of this 

 method is unlimited and it is admirably adapted to clusters of stars 

 like oL Centauri in which 128 variables were discovered at Arequipa. 

 At Harvard this ingenious method has led to the discovery of 2,000 

 to 3,000 variables, several times as many as have been made know^n 

 by all the other methods combined. 



The objective prism aids also in this class of discovery. We Imow, 

 for instance, that the variables of long period, of the type of Mira 

 Ceti, for example, contain bright lines in their spectra ; on the other 

 hand, these bright-line stars are comparatively rare, so that a star 

 possessing this characteristic attracts our attention at once to its 

 variability. 



The organization of this bureau, composed solely of women and 

 which has executed such colossal work, merits our attention. The 

 reader may, however, reassure himself; we shall not detain him long. 



