the mathematical process by which M. Hansen arrived at this result, but there is no difficulty in understanding the general principles on which it is founded. In discharging a ball from a gun, calculation can predict che trajectory it will describe. But if the ball is not equally dense on opposite sides, it will not pursue the same path it would do if homogeneous. Let us suppose, that while the ball is perfectly spherical, one half is iron and the other cork, the curve described will be different, both in range and form, from that which would be described by a ball equally dense throughout. Balls have been, indeed, purposely so cast, to increase the range—the sphere being hollow, but having one side thicker than the other. Given the difference of density, the curve can be laid down, and given the curve, the difference of density can be determined. This last case is that of the moon. It differs in no respect from a ball discharged from a gun, and, in examining the curve it describes, the conclusion is, that while she is quite or nearly spherical, the hemisphere, turned towards us, is lighter than the opposite one.
But how does this tell on the question of inhabitants? The application is very direct and startling. Supposing the sphere of the moon originally covered with water, and enveloped in an atmosphere, both water and air would flow to the heavier side, and leave the lighter side destitute of both, just as water and air leave the summits of our mountains, and gravitate
