surface. In a somewhat wider zone, the planetary structure could be only preserved under the form of an ellipsoid, which would deviate widely from a sphere, not so much on account of its rotation as from the effects of the unequal attraction which its different parts receive from the primary. The accompanying diagram shows the equilibrium form of a satellite moving in a circle around a large sphere, both bodies
being equally dense and separated by small distances. In a somewhat greater proximity to the primary, the satellite would be reduced to a state of instability; and its matter, if not kept together by a great cohesive power, would scatter into independent orbits, and ultimately form a ring. There are, perhaps, in our solar domain too few cases to show how excessive tidal action can produce its definite results in its various degrees of power. If the nearest satellite of Mars bore to its primary such relations of size and density as subsist between the earth and the moon, it would, no doubt, surpass all known planets in deviating from a true sphere, even though its materials were not of the most yielding character. If similar relations existed between Saturn and his closest secondary, the latter would show a greater deviation than the primary does from a spherical form.