MIRACLES OF SCIENCE 



will come within the danger line of Roche's limit, or 

 actually plunge together. 



In either case the stars must explode like gigantic 

 bombs, and it seems unlikely that the resulting neb- 

 ular masses would take on the regular form of the 

 nucleated spiral. We should rather expect the gase- 

 ous mass to assume a more or less globular form in 

 case of explosion without actual contact; a cone- 

 shaped mass if the two stars came together at a con- 

 verging angle ; or a long drawn out cloud in the event 

 of a glancing head-on collision. 



And in point of fact the telescope fully justifies these 

 expectations. There are multitudes of nebulae that do 

 assume forms widely different from the typical spiral. 



There is a famous nebula in the constellation Sagi- 

 tarrius, for example, which is called the trifid nebula 

 because it seems to be split into three somewhat reg- 

 ular parts. Again there is the network nebula in the 

 constellation Cygnus, which is drawn out into a long 

 irregular curiously cloud-like streak of nebulosity; 

 and there are sundry other nebulae that are disc-like 

 in form, and a few that have the shape of rings, 

 notably one in the constellation Lyra which is visible 

 through a small hand glass. 



We cannot suppose that nebulae of these aberrent 

 types will develop into solar systems such as ours. 

 They seem destined to produce clusters of stars rather 

 than a single star with small planetary attendants. 



AN ALTERNATIVE THEORY OF THE FORMATION OF NEBULAE 



Although the theory of collisions between stars 

 gives a rather plausible explanation of the origin of 



20 



