3io] Conclusion 289 



tenuous gas in slow rotation. As imagined by Laplace, this mass contracted 

 owing to loss of energy by radiation, and so increased its angular velocity 

 until it assumed a lenticular shape, similar to that of the figures shewn on 

 Plate II. After this, further contraction was a sheer mathematical impossi- 

 bility and the system had to expand. The mechanism of expansion was provided 

 by matter being thrown off from the sharp edge of the lenticular figure, the 

 lenticular centre now forming the nucleus, and the thrown-off matter forming 

 the arms, of a spiral nebula of the normal type. The long filaments of matter 

 which constituted the arms, being gravitationally unstable, first formed into 

 chains of condensations about nuclei, and ultimately formed detached masses of 

 gas. With continued shrinkage, the temperature of these masses increased 

 until they attained to incandescence and shone as luminous stars. At the same 

 time their velocity of rotation increased until a large proportion of them 

 broke up by fission into binary systems. The majority of the stars broke 

 away from their neighbours and so formed a cluster of irregularly moving 

 stars our present galactic universe, in which the flattened shape of the 

 original nebula may still be traced in the concentration about the galactic 

 plane, while the original motion along the nebular arms still persists in the 

 form of "star-streaming." In some cases a pair or small group of stars failed 

 to get clear of one another's gravitational attractions and remain describing 

 orbits about one another as wide binaries or multiple stars. The stars 

 which were formed last, the present .B-type stars, have been unusually im- 

 mune from disturbance by their neighbours, partly because they were born 

 when adjacent stars had almost ceased to interfere with one another, partly 

 because their exceptionally large mass minimised the effect of such inter- 

 ference as may have occurred ; consequently they remain moving in the 

 plane in which they were formed, many of them still constituting closely 

 associated groups of stars the moving star- clusters. 



At intervals it must have happened that two stars passed relatively neai 

 to one another in their motion through the universe. We conjecture that 

 something like 300 million years ago our sun experienced an encounter of 

 this kind, a larger star passing within a distance of about the sun's diameter 

 from its surface. The effect of this, as we have seen, would be the ejection 

 of a stream of gas towards the passing star. At this epoch the sun is sup- 

 posed to have been dark and cold, its density being so low that its radius 

 was perhaps comparable with the present radius of Neptune's orbit. The 

 ejected stream of matter, becoming still colder by radiation, may have con- 

 densed into liquid near its ends and perhaps partially also near its middle. 

 Such a jet of matter w.ould be longitudinally unstable and would condense 

 into detached nuclei which would ultimately form planets. The more liquid 

 planets at the end of the chain would be those of smallest mass ; the gaseous 

 centre would form the larger planets Jupiter and Saturn. Owing to the 

 orbital velocity which had been communicated to these planets by the 

 -i.e. 19 



