316 CARNEGIE INSTITUTION OF WASHINGTON. 



in the same form as before, their mutual backward pull will reduce 

 their velocities approximately to their former velocities, save for some 

 interchange due to partition of energy. The algebraic sum of their 

 moments of momentum must be preserved, but as this algebraic sum 

 is the difference of the momenta of the two stars, it is consistent with 

 a wide range of individual velocities offsetting one another. The 

 critical question, therefore, lies in the effect which the deployment due 

 to the eruptive projection of some large part of the star-masses will 

 have upon the velocities of the newly formed nebulae as they, in place 

 of the stars, retire from one another. According as there may be any 

 degree of reduction of the backward pull, an enhanced velocity, com- 

 pared with the normal star velocity, will be retained. 



The simplest line of inspection may be made clear by an appeal to 

 general principles, as follows: (1) In so far as the stars remain spherical 

 during the whole history of the interaction, no effect (other than the 

 normal partition of energy) will follow ; (2) in so far as elongations and 

 projections are developed, and these are directed in the line of mutual 

 attraction of the stars, or dominantly referable to it, the gravity effect 

 will be increased; (3) in so far as the axis of such elongations and pro- 

 jections is dominantly transverse to the line of mutual attraction, the 

 gravity effect will be reduced. 



Now in all the earlier stages of approach, the Hues of elongation and 

 proj ection will diverge from the lines of mutual attraction by a small angle 

 only, and hence tidal elongation and eruptive proj ection will increase the 

 velocities of the approaching stars. The early stages of the assigned 

 deployment thus add to the high velocities of approach against which 

 drafts are to be made later by backward pull after periastron is passed. 



But near periastron the courses of the stars are sharply curved 

 into directions largely transverse to their pre\dous courses and the 

 gravity effect of elongation and projection is reversed. New projec- 

 tions should continue to take place until periastron is passed, unless 

 the expansiveness of the stars has been exhausted previously. These 

 projections will, like the previous ones, be directed, as they start forth, 

 in the changed line of attraction, but they will quickly be brought into 

 a transverse attitude by the swift motion of the stars at this climacteric 

 stage. Some lag in the eruptions at all stages must be presumed, so 

 that the transverse attitudes will be delayed and will tend to fall mainly 

 after periastron is reached — i. e., they will fall in the stages in which 

 mutual attraction retards the previously acquired velocities. In so 

 far as they thus reduce the backward pull they leave the nebulse, in 

 passing on, a larger portion of the velocities previously attained. 



After the stars have swung about one another and begin to pass 

 away in nearly opposite directions, their attitudes toward the projec- 

 tions are again reversed except as rotation or revolution has intervened 

 to change these. While the rotation or revolution of the projected 

 matter greatly complicates the case, it appears that each couplet of 



