mics. We are placed at too great a distance from the small planets or asteroids, as they are called, to be able to see whether or not they have any gaseous surroundings. But it is possible, from the ingenious argument we are discussing, to assure ourselves that such small bodies must be quite as devoid of air as the moon. There are, we know, globes in our system only a few miles in diameter, and so small in mass, that a cricket-ball there, receiving the velocity it would get from the bat of a W. G. Grace, would go off into space never to return. It is quite obvious that the molecules of any gases we know would be far too nimble in their movements to remain prisoners at the surface of little globes of this description, to which their only bond was the feeble attraction of gravitation. It is, therefore, in the highest degree improbable we might, indeed, almost say impossible for gaseous surroundings to be preserved by any globe where the mass is not considerably greater than that of the moon.
In applying these considerations to Mars we have first to note that its mass and size are intermediate between those of the earth and the moon. It is much more capable of retaining an atmosphere than the moon, though its capability in this respect falls short of that possessed by the earth. But in such a case it is essential to depend not on mere generalities but on the actual numerical facts of the case. Without going too deeply into detail it is sufficient to observe that there must be for each globe a certain critical velocity represented by the least pace at which a missile projected from it will succeed in escaping altogether. In discussing this we may leave out of view the question of the resistance which the air opposes to
