130 ANNUAL REPOET SMITHSONIAN INSTITUTION, 1932 



The question of the origin of eclipsing and spectroscopic binaries, 

 according to some astronomers, has a close connection with the ques- 

 tion of interrelationships of the different classes of variable stars : 

 Jeans has proposed the theory that all variable stars are rotating and 

 that when the period of rotation becomes sufficiently rapid, as in the 

 case of the short-period Cepheids, the stars can no longer hold 

 themselves together — a hypothesis known as the fission theory. If 

 then a Cepheid is the parent of a binary system, how are we to prove 

 it? Astronomical questions rarely lend themselves to direct proof, 

 unfortunately, but any clue serves as a valuable bit of evidence either 

 for or against a theory in question. We have evidence for extending 

 the connecting links between long-period and Cepheid variables to 

 eclipsing and spectroscopic binaries as well. We may mention: (1) 

 There are spectroscopic evidences indicating rotation of Cepheids 

 about their respective centers of gravity; (2) the mean densities of 

 Cepheids are approximately what we should expect of two embryo 

 stars in actual contact; (3) the mean radii of Cepheids compare 

 favorably with the separation of the components of short-period 

 binary systems; (4) slight evidences of correlations of colors and 

 brightnesses of short-period binaries with Cepheids are to be found ; 

 (5) the agreement of periods is as good as can be expected; and (6) 

 the space distributions of the two classes of objects compare favor- 

 ably. 



Of these arguments the first must carry considerable weight. B}'^ 

 means of the spectroscope it has been found that motion of the star 

 or of some part of it toward the observer is greatest at maximum 

 light or shortly thereafter. It is also believed from the evidences 

 now available that the star is hottest at maximum light, approxi- 

 mately. Both can not be true of a pulsating star; mathematical 

 analysis involving considerations of luminosity, radius, and tem^ 

 perature shows that the star should be neither expanding nor con- 

 tracting, or in other words, it shows that the velocity in the line of 

 sight should be zero at maximum light, provided that the tempera- 

 ture is hottest at about this time, as is generally believed. On the 

 other hand, however, either of the two systems in Figure 1 should 

 show greatest velocity of approach at maximum light if maximum 

 light is to be associated with maximum visible surface. To agree 

 with observations, the part of the Cepheid from which most of the 

 light apparently comes at this moment should also be hottest, and 

 consequently a very unequal temperature distribution over its sur- 

 face is demanded on the basis of the fission theory. It is evident 

 therefore, that although most of the above six arguments are neces- 

 sary conditions to substantiate a fission theory such as that of Jeans, 

 they are not sufficient; they prove nothing. 



