Bickrrton,—0n Temporary and Variable Stars, 123 
whilst, on the other hand, we have inertia and gravity tending to keep the 
mass together. The centrifugal force acts only in one plane, whilst the re- 
pellent force of heat acts in every plane; a bun-shaped mass must result, 
with a number of distinct pieces, which at first at least are in advance of the 
general mass. Follow it on in time and we get the ring nebula, with or 
without a luminous centre; in the latter case, with a dark circle dividing 
those parts whose velocity has carried them beyond the powers of the 
attractive force, from those parts held prisoner by it, These parts, as they 
gradually radiate heat into space, are once more slowly attracted to the 
centre by gravitation. If the piece struck off from each body were very 
small, then complete dissipation of the whole into space would result, 
Clearly such collisions as I have described would be competent to produce 
every variety of temporary stars that has appeared. Applying the spectro- 
scope to such a star, we get at first a continuous spectrum; then black 
lines, quickly followed by bright lines and spectrum; then bright lines 
alone. Again, if the colliding bodies were of very different size, or if the 
heat were not great enough to entirely volatilize the star, we should have 
lines and spectrum. Lastly, as heat and pressure diminish by the dissipa- 
tion of the body into space, we get fewer and fewer lines, until only those 
substances in greatest quantity, or of greatest power in giving lines at 
lowest temperature and pressure, remain luminous, and we have a nebule 
left; or in the case of total dissipation of the gaseous mass all evidence 
of its existence will disappear. It will be seen how exactly the above 
hypothesis agrees with the spectroscopic observation of temporary stars; 
and I have showa as fully as perhaps it is wise to do in this paper, that the 
hypothesis of partial impact is competent to account for every variety of 
these bodies, and also for their intensity and short duration. 
We must now return to the parent bodies which we left travelling on in 
space. A cylindrical or curved slice has been cut out of each ; sometimes 
that is the chief thing that will happen. But on the other hand we may have 
the molten interior of the body exposed to view. If there were atmospheres 
on the two colliding bodies, a very great heating of the surface of the section 
would result, and when both causes are acting in unison a stupendous lake 
of fire must be formed. Let such a body rotate on its axis, alternately the 
light and dark sides are shown, and we get a variable star. May not Mira in 
this way be attempting to tell us her autobiography ; how she is a dark body, 
with a molten lake of fire, 80 degrees of are, a lake as big as our sun, and 
how she rotates about an axis in a little less than a year? If it be so, she 
tells us of a dark body almost as large as Sirius, or how would 30 degrees moe 
of are produce a star of the first magnitude? Algol appears to tell us that oo 
it is a dark and gloomy PM with a brilliant son who were! passes 
