80 EEPOKT— 1891. 



the mass of the body not having sufficient attractive power to 

 retain them. 



8. Hence an intensely-briUiant body is produced in less 

 than an hour ; it then expands, and increases in size and total 

 luminosity for perhaps a few hours to a day or so ; then 

 the diffusion would be so great as to gradually lessen lummo- 

 sity, until in a few months or a year the star would have 

 disappeared into space. This represents all the peculiarities 

 of temporary stars. 



9. If the graze be more considerable the attraction will be 

 greater, yet the molecular velocity the same : a hollow globe of 

 gas may then result, giving us a planetary nebula. Accord- 

 ing to Lord Lindsay this is the condition of the temporary star 

 of the Swan. 



10. If the graze be still more considerable the slower- 

 moving molecules of heavy metals may not travel far from the 

 centre, and may be attracted back and form a brilliant mass, 

 giving us the appearance of a nebulous star. 



11. The two stars that grazed each other will be heated on 

 the colliding sides.* They will also be set revolving in such a 

 manner that the heated sides are alternately presented on 

 opposite sides of the heavens, giving us the appearance of a 

 pair of variable stars. But this inequality of temperature 

 would slowly subside, but at different rates in the two stars, 

 leaving only one variable. 



12. If this w^ere so, then variable stars should be sometimes 

 in pairs. The 100 variables in Chambers's list were charted to 

 ascertain if this was the case : forty were found to be in pairs, 

 twenty so close as to occupy only ten spaces ; they were too 

 close not to coalesce when drawn on a lOin. chart, a result 

 absolutely free from the possibility of being a chance coin- 

 cidence. 



13. If variable stars were formed this way they would 

 lessen iii irregularity, and also would for many reasons have 

 inequality of times of rotation and inequality of intensity. All 

 these peculiarities are known to be common to variable stars. 



14. If double stars were formed this way they would some- 

 times be variable. After three years of search for this fact, it 

 w^as discovered (in 1862) that Struve has found this property 

 to be common with double stars. They are often coloured — a 

 result that must generally follow and be associated with varia- 

 bility in consequence of the heavy metallic centre being brought 

 to the surface. 



15. If this be the origin of double stars, they should also 

 often be associated with nebulae. Herschel says, " The asso- 

 ciation of double stars and nebulae is truly remarkable." 



16. If two bodies were composed of hydrogen and were to 

 collide, the temperature would be two hundred times less than 



