248 The Evolution of Binary and Multiple Stars [OH. xi 



progresses the atmosphere will continually increase at the expense of the 

 central mass. This motion is one which it is exceedingly difficult to trace out 

 dynamically, even in its main outlines. It seems possible that the atmo- 

 sphere may in time condense into nuclei, and that these might ultimately 

 form planets, but this has to be mainly a matter of conjecture. It is perhaps 

 worth considering whether the planetary, and possibly also the ring nebulae, 

 can be bodies of the type we have been considering*. We may also con- 

 template the possibility of planets being formed in this way although, as we 

 shall see later, the planets of our solar system cannot have been so formed. 



258. The equatorial ejection of matter will continue until a further 

 critical density p is reached, at which the pseudo-spheroidal figure for the 

 nucleus becomes unstable and gives place to a pseudo-ellipsoidal form 

 (cf. 185). The new pseudo-ellipsoidal figure will eject matter only at its 

 two pointed ends ; it is perhaps worth considering spindle-shaped planetary 

 nebulae such as N. G. C. 7009 in this connection (see Plate I). Ultimately 

 the pseudo-ellipsoidal nucleus gives place to a pear-shaped figure and this 

 will divide into two detached masses. 



The final result of the process of disintegration will accordingly be a binary 

 star, the two components rotating about one another in a more or less dense 

 atmosphere of ejected matter, through which they will plough their way. 

 This formation at once recalls Duncan's attempted explanation of the Cepheid 

 variables f. At a later stage the atmospheres will condense round the two 

 stars, leaving an ordinary binary star. 



During the process of condensation, the greater part of the atmosphere is 

 likely to condense round the more massive constituent, so that the light from 

 the more massive star will be more screened than that from the lighter one. 

 Temporarily the more massive component may shine less brightly than its 

 smaller companion. This condition is observed in ft Lyrae, a spectroscopic 

 binary in which the dark star has a mass about 2 '2 times that of the bright 

 star, the two masses being very nearly in contact ; the explanation we have 

 given of this condition was first suggested by Meyers J. If this is the true 

 explanation, it would appear that @ Lyrae provides an instance of a binary 

 star in the very earliest stages of its existence. 



259. If the stars may be regarded as masses of ordinary gas, it is not 

 difficult to obtain an estimate of the critical density p before which fission 

 cannot begin. 



Discussing a mass of gas in adiabatic equilibrium, we found in Chapter VII 

 that fission could not begin so long as 7, the ratio of the specific heats, was 

 less than about 2-2. 



- See Plate I (p. 5) ; also Campbell, Science, 45 (1917), p. 538. 



f Lick Obs. Bull. 6 (1911), 151. See also Campbell, Stellar Motions, p. 307. 



J Astrophys. Journ. 7 (1898), p. 21. 



