RECENT ADVANCES IN SCIENCE 19 



up first, and that the distance apart of its two components will 

 be small compared with that between the original pair. Some 

 of the many binary and multiple systems in the sky appear 

 to show these characteristics, and it is possible that they were 

 formed in this way, although Jeans has recently concluded 

 (M.N., R.A.S. 79, April 1919), from a statistical study of the 

 distribution of eccentricities and periods of binary systems, that 

 the probability is against these systems having been so formed. 



The conditions postulated in these researches are doubtless 

 very far removed from reality, and they do not therefore pro- 

 vide a criterion as to the validity or otherwise of the nebular 

 hypothesis. In 1855 Roche discussed the other extreme 

 limiting case in which the matter is so compressible that it is 

 practically concentrated close to its centre of gravity, the outer 

 parts being merely of the nature of an " atmosphere " of small 

 density. In this case it is found that, after a time, a lens- 

 shaped figure is formed, which develops a sharp edge round 

 its equator, from which, after a certain critical velocity has 

 been attained, any further increase in velocity results in the 

 ejection of matter from the periphery. The behaviour of such 

 a mass is therefore very similar to what Laplace predicted, but 

 again we are faced with the fact that the case is an extreme 

 one. 



Jeans has now shown, in some brilliant mathematical 

 investigations, that in intermediate cases a compressible 

 mass will behave similarly to Roche's model, provided that 

 its density is less than about a quarter of that of water, but 

 that above this limiting density there is a sudden change, 

 and the behaviour approximates to that of an incompressible 

 mass. He concludes, therefore, that young masses in rotation 

 break up in the way pictured by Laplace, and old masses in the 

 way pictured by Darwin. 



In this division into young and old masses, the majority 

 of multiple systems observed in the sky would fall into the 

 latter category, having masses above a quarter of that of 

 water. Their behaviour is therefore in accordance with 

 Jeans 's conclusions. 



Jeans considers that the other class — rotating bodies of 



•small density throwing off matter from their equators — are 



represented by the spiral nebulae. Photographs of spiral 



nebulae which are edgewise-on to us show that the central 



