Periodical comets are evidently bodies which have each lived through a chapter of accidents, and a significant hint as to the nature of their adventures can be gathered from the fact that their aphelia are pretty closely grouped about the tracks of the major planets. Halley's, and five other comets are thus related to Neptune; three connect themselves with Uranus, two with Saturn, above a score with Jupiter. Some form of dependence is plainly indicated, and the researches of Tisserand,[265] Callandreau,[266] and Newton[267] of Yale College, leave scarcely a doubt that the "capture-theory" represents the essential truth in the matter. The original parabolic paths of these comets were then changed into ellipses by the backward pull of a planet, whose sphere of attraction they chanced to enter when approaching the sun from outer space. Moreover, since a body thus affected should necessarily return at each revolution to the scene of encounter, the same process of retardation may, in some cases, have been repeated many times, until the more restricted cometary orbits were reduced to their present dimensions. The prevalence, too, among periodical comets, of direct motion, is shown to be inevitable by M. Callandreau's demonstration that those travelling in a retrograde direction would, by planetary action, be thrown outside the probable range of terrestrial observation. The scarcity of hyperbolic comets can be similarly explained. They would be created whenever the attractive influence of the disturbing planet was exerted in a forward or accelerative sense, but could come only by a rare exception to our notice. The inner planets, including the earth, have also unquestionably played their parts in modifying cometary orbits; and Mr. Plummer suggests, with some show of reason, that the capture of Encke's comet may be a feat due to Mercury.[268]