2t8 CARNEGIE INSTITUTION OF WASHINGTON. 



altogether amount to about one seven-hundredth part of the mass 

 of the system. Simply to supply the required planetary matter, the 

 protuberances need include but this small fraction of the ancestral 

 sun. However, some considerable part of the projected matter 

 must probably have been gathered back into the sun, and some part 

 may possibly have been projected beyond the control of the system. 

 Making allowances for both the.se factors, the proportion of the 

 sun's mass necessarily involved in the protuberances is still very small. 

 Apparently i or 2 per cent of the sun's mass would amply suffice. 



The protuberances, by hypothesis, would be thrust out as the sun 

 was swinging about its temporary companion star in a sharp curve, 

 and necessarily at a prodigious velocity. It is inferred that the pro- 

 jected protuberances would be differentially affected by the attraction 

 of the companion star, and take different curves about it, out of 

 which must spring rotatory motion. This seems logically clear, but 

 the precise paths pursued by the parts of the protuberances would 

 apparently vary widely with different cases. As each case consti- 

 tutes an involved example of the problem of three bodies, the whole 

 is beyond rigorous mathematical treatment, but special solutions seem 

 to justify the inference that effective rotation would arise. 



The distal portions of the protuberances would obviously be formed 

 from the superficial portions of the sun, while the later portions of 

 the ejections forming the proximal parts of the arms would doubtless 

 come mainly from lower depths, and hence probably contain more 

 molecules of high specific gravity. In this seems to lie a better 

 basis for explaining the extraordinary lightness of the outer planets 

 and the high specific gravities of the inner ones, than in the separa- 

 tion, from the extreme equatorial surface of a gaseous spheroid, of 

 successive rings whose total mass only equaled one seven-hundredth 

 part of the original nebula. 



It seems consistent with the conditions of the case to assume that 

 the protuberances would consist of a succession of more or less irreg- 

 ular outbursts, as the ancestral sun in its swift whirl around the 

 controlling star was more and more affected by the latter' s differen- 

 tial attraction ; and hence the protuberances would be directed in 

 somewhat changing courses, and would be pulsatory in character, 

 resulting in rather irregular and somewhat divided arms, and in a 

 knotty distribution of the ejected matter along the arms. These 

 knots must probably be more or less rotatory from inequalities of 

 projection. 



