Page:Astronomy for Everybody.djvu/321

may be of the nature of bubbles, but this is far from being established.

A star, like the sun, must be hotter in the interior than at its surface. From the latter alone can heat be radiated; hence the surface is continually cooling off, and if the matter composing the body were at rest, the cooling would soon go so far that a crust would form, as it does on a mass of molten iron. The only way in which this can be prevented is that, as the superficial portions cool, the greater density which they thus acquire causes them to sink down into the seething mass below, portions of which arise to take their place, cool off, and sink in their turn. Thus there is a continual interchange of matter between the inside and the surface, much as in a boiling pot the water at the bottom is continually being forced up to the top, while that on the top continually sinks down.

It follows from this that there must be a limit to the smallness of a star. If such a body were no larger than the moon, it would, in a few thousand years, so far cool off that a crust would form over its surface. This would cut off the currents by which the hot matter is brought to the surface and the star would soon cease to shine. As there can be little doubt that the age of most of the stars is to be reckoned by millions of years, it follows that they must be so large that they can lose heat for millions of years and yet a cool crust not form on their surface.

We have said that our sun is among the colder of the stars and also that it is among the smaller. These two facts fit well together. The smaller a star is the more