i92 Transactions. — Miscellaneous, 



iron, only on its velocity. So when our dead suns come into impact, t1a6 

 temperature will not depend on the size of the spark cut off, only on the 

 velocity destroyed. But if the piece cut off be small, it has but little attrac- 

 tive power to keep it together, and the particles are so hot — or moving so 

 fast — that every single molecule flies completely away and disappears into 

 space. Does it not look extremely likely that here we have our temporary 

 stars, bursting forth when the collision occurs, and disappearing when its 

 particles travel away into space ? It certainly appears very likely. 



But what about the two large pieces (the two wounded stars) ; a slice 

 has been cut off each, and the hotter interior exposed ; friction has also 

 developed heat, and so when they become round again they will be hotter 

 on one side than on the other. As they revolve they must almost certainly 

 form a variable star, and the struggle of the two rotations will make this 

 variation pass through long cycles, just as a spinning top oscillates if it has 

 a kick. But as it would seem that two variable stars must be often pro- 

 duced together, the lists were searched to see if any pairs could be found, 

 and a chart has been made from Chambers' hst, and it shows sixteen well- 

 marked pairs, or thirty-two connected stars out of one hundred and twenty. 

 Unless we suppose this spotted condition to be a disease and catching, it is 

 difficult, except on this view, to account for the pairs. Not only do we 

 thus find these pairs existing, but some variable stars are close to the places 

 where old temporary stars formerly existed, and also variable stars have 

 become ordinary stars, as we should expect them to do when the tem- 

 perature became uniform ; and doubtless when the whole are carefully 

 matched, many will be found to be gradually approaching the state of uni- 

 formity exhibited by ordinary stars. But it is not necessary to suppose 

 thate th piece struck off each should always be such an excessively small 

 ratio as to be projected into space, although the temperature produced by 

 the collision will be almost always high enough to make gas of the coalesced 

 part. This part may have mass enough to remain a j)ermanent nebula. 

 In this case reasons have been urged that render it probable that at first 

 this gas would tend to take a spindle shape. Afterwards many possibilities 

 present themselves according to the varying circumstances of the collision, 

 for it is perfectly evident that these may be very numerous indeed. As the 

 bodies may vary from cold dense sohd bodies to rare masses of hot diffused 

 gas, they may origmally be moving very fast or very slowly ; they may have 

 been spinning with great velocity or hardly rotating at all ; they may be 

 nearly the same size, or a very unequal pair ; and it does not need a Newton 

 to see that any of these states will influence the result attained at collision 

 and afterwards. The only effects which appear absolutely certain to follow 

 partial impact are that rotation must ensue, that the matter will tend to 



