with a gradual augmentation of velocity at the rate we have named. Had our atmosphere been originally in a highly diffused state, gravitation would have drawn it in to the stable condition which it at present occupies. The condition of equilibrium in our atmosphere appears to be as follows. At the surface of the earth there is of course an unyielding surface so far as the air is concerned. We may regard the atmosphere as divided into a number of concentric shells around the solid sphere. Each shell has a density less than that of the shell above it, and greater than that of the shell below it. The upward pressure from the lower shell compensates for the effect of gravitation on the shell above, and thus the equilibrium is sustained. But in the case of the sun there seems to be no solid shell possible. The consequence is that there must be a general shrinkage of the entire mass; this being so the molecules on the whole get nearer to the sun's centre, and consequently, in virtue of the attractive power of the sun, work is done on them and they acquire enhanced velocity. Thus on the whole the velocities of the solar molecules in consequence of the solar attraction tend to increase.
On the other hand there is a distinct loss of energy to which the molecules are exposed. As the velocities increase the encounters between the different molecules become more severe. After each such encounter the molecule vibrates with increased energy, obtained at the expense of the velocity of translation. The molecules are thus rendered more competent to impart energy to the surrounding ether, that is to say, they acquire an increased power of radiation. Thus we see that the fact of the sun's contraction translates itself with a tendency
