the passage of the projectile. This is, no doubt, of vital importance in cases where actual artillery practice is concerned, yet it is not material to our present inquiry. The problem which we are considering depends on the movements of the molecules of air at the summit of the atmosphere, and the question is simply whether after they have made an incursion into free space there is sufficient efficiency in the attraction of the globe to effect their recall.
At the surface of Mars the speed which would carry a body away from its surface altogether is about three miles per second. It seems certain that the velocity of the molecules of hydrogen is often far in excess of this, and consequently free hydrogen is impossible as a permanent ingredient of the Martian atmosphere. Oxygen, however, has a molecular velocity only about one-fourth of that of hydrogen, and it seems unlikely that the oxygen molecules can ever have sufficient velocity to permit their escape from an atmosphere surrounding Mars. There is nothing, therefore, to prevent this element from being now present.
But the case of the vapour of water is especially instructive and interesting. Its molecules have a speed which averages about one-third of that attained by the molecules of hydrogen. It would seem that the utmost pace that the molecules of water could attain (being perhaps seven times the average velocity) would be about 2 miles per second. Now this would not be enough for escape from Mars, for we have seen that a speed of three miles per second would be required for this purpose. This argument suggests that the globe of Mars happens to approach very closely the dimensions and mass of the
