contain the least. In consequence of this principle, the atmosphere of the smaller body finds itself obliged to cover relatively more surface, which still further thins it out.

Lastly, gravity being less on the surface of the smaller body, the atmosphere is less compressed, and, being a gas, seizes that opportunity to spread out to a greater height, which renders it still less dense at the planet's surface.

Thus, for three reasons, Mars should have a thinner air at his surface than is found on the surface of the Earth.

Calculating the effect of the above causes numerically we find that on this a priori supposition Mars would have at his surface an atmosphere of about fourteen hundredths, or one seventh, of the density of our terrestrial one.

Observation supports this general supposition; for the cloudless character of the Martian skies is precisely what we should look for in rare air. Clouds are congeries of globules of water or particles of ice buoyed up by the air about them. The smaller these are, the more easily are they buoyed up, because gravity, which tends to pull them down, acts upon their mass, while the resistance they offer varies as the surface they present to the air, and this relatively greater in the smaller particles. The result is that the smaller particles can float in thinner air. We see the principle exemplified in our terrestrial clouds; the low nimbus being formed of comparatively large globules, while the high cirrus is made up of very minute particles. If we go yet higher, we reach a region incapable of supporting clouds of any kind, so rarefied is its air. This occurs about five miles above the Earth's surface; and yet even at this height the density of our air is greater than is the probable density of the air at the surface of Mars. We see, therefore, that the Martian atmosphere should from its rarity prove cloudless, just as we observe it to be.

So far in this our investigation of the Martian atmosphere we have been indebted solely to the principles of mathematics and molar physics for help, and these have told us something about the probable quantity of that atmosphere, though silent as to its possible quality. On this latter point, however, molecular physics turns out to have something to say; for an Irish gentleman, Dr. G. Johnstone Stoney, has recently made an ingenious deduction from the kinetic theory of gases bearing upon the atmospheric envelope which any planet can retain. His deduction is as acute as it appears from observation to be in keeping with the facts. It is this : --

The molecular theory of gases supposes them to be made up of myriads of molecules in incessant motion. What a molecule may be nobody knows; some scientists supposing it to be a vortex ring in miniature,-- something like the swirl made by a teaspoon drawn through a cup of tea. But, whatever it be, the idea of it accounts very creditably for the facts. The motion of the molecules is almost inconceivably swift as they dart hither and thither throughout the space occupied by the gas, and their speed differs for different gases. From the observed relations of the volumes and weights of gases to the pressures to which they are subjected is deduced the fact of this speed and its amount. It appears that the molecules of oxygen travel, on the average, at the rate of fifteen miles a minute; and those of hydrogen which are the fastest known, at the enormous speed of more than a mile a second. But this average velocity may, for any particular molecule, be increased by collisions with its neighbors. The maximum speed it may thus attain Clerk-Maxwell deduced from the doctrine of chances to be sevenfold the average. What may thus happen to one, must eventually happen to all. Sooner or later, on the doctrine of chances, each molecule of the gas is bound to attain this maximum velocity of its kind. When it is attained, the molecule of oxygen travels at the rate of one and eight tenths miles a second, the molecule of water vapor at the rate of two and one half miles a second, and the molecule of hydrogen at over seven miles a second, or four hundred and fifty times as fast as our fastest express train.

Now, if a body, whether it be a molecule or a cannon-ball, be projected away from the Earth's surface, the Earth will at once try to pull it down again: this instinctive holding on of Mother Earth to what she has we call gravity. In the cases with which we are personally familiar, her endeavor is eminently successful,