directions. This rule will evidently need but slight modifications, when, instead of being so extremely unequal, both bodies have the same relation of size as that subsisting between our earth and moon, or even such as is represented in our diagram. In this case, however, the lesser body would bear a somewhat greater disturbing influence; but its dismemberment, though of a paroxysmal and a very extensive character, would be confined to the side next the primary. On losing a large portion of its mass, the satellite would swing into a wider orbit; its distance from the primary would for a long period be increased by tidal action, and many ages must elapse before they again became near enough to give occasion for a like convulsive rupture. The incorporation of a large celestial body with a greater one around which it previously revolved would thus be effected by a number of paroxysms, and would not be completed before many billions of years.
The intermittent character of these rare events would be very decided, except, perhaps, when the subordinate body were, like a comet, composed of a profusion of exceedingly rarefied gas surrounding a small, dense, central nucleus. Such differences of density as may be naturally expected in the internal and superficial matter of a satellite would tend to give the convulsive dismemberment a somewhat reduced scale, and to make it recur after less remote periods of time. But this influence would be more than counterbalanced if the incorporating body were solid, as the planetary structure would be preserved for a longer time; but, when the rupture took place, the ruin would be more extensive. Indeed, in the cases most likely to occur, the doomed planet would meet its fate in successive stages, of which the number and magnitude may be estimated with tolerable accuracy. If our moon were made to revolve about 4,500 miles beyond our atmosphere, its coalescence with our globe would be inevitable, and it would take place by about six or eight paroxysmal stages extending over a vast immensity of time. Two or three times the number of such terrific convulsions may be expected in the union between Algol and the large planet which causes his variability; and the same estimate will serve for the binary systems or the physically double stars when after long ages they become close enough for the incorporation of the less with the greater.
It is the terrific conflict of matter on such rare and stupendous events, that awakens the power which is mainly concerned in giving birth to worlds. Large primary planets would be called into being if one or both of the celestial objects undergoing these violent stages of combination had the rank of a sun. The vast mass of matter precipitated to the greater body on these occasions would sweep along its equator with furious velocity. But on the subordinate one, especially in its equatorial regions, the more superficial parts would slide over the internal nucleus in an opposite direction, in consequence of the tidal action, which in the new orbit must be powerful enough to produce not merely waves, but even progressive movements at the rate of many