The impact theory seems to remove all these difficulties. It is just as likely _à priori_, if not more so, that the primitive form of the protyle should have been that of large cold masses moving through space in all directions, with excessive velocities, as that it should have been that of a gaseous mass in a state of unstable equilibrium. If we assume the former condition, then the colliding of these masses would account not only for the ultra-gaseous state, but also for its inconceivably high temperature. Besides, in this case we are not called upon to account for any other antecedent state of the masses before collision, for they may have existed from the beginning of creation in the form of masses in motion through space. Had space and time permitted, it might have been shown that there are other obscure points on which the theory seems to shed additional light. I shall now, in conclusion, refer to a point wherein the theory differs radically from that of all other theories of stellar evolution. But before doing so I may briefly refer to an objection which has been frequently urged against the theory. _Objection considered._—The objection to which I refer is this, that, had the nebulæ been produced by impact in the way implied in the theory, then we ought to have had some historical record of such an event. I can perceive no force in such an objection. Our historical records, I presume, do not extend much farther back than about 3,000 years, and we have no evidence to conclude that a new nebula makes its appearance in the visible firmament with such frequency; and supposing it did, we have no grounds for assuming that its production by impact in the way supposed by the theory would attract general notice. It is doubtful if the nebula produced would, in the first instance, be actually visible. I have shown that the temperature of the nebula could not have been less than about 300,000,000° C., and it is very doubtful if the gaseous mass enveloping all that was solid in the nebula would, at such a temperature, be self-luminous. The probability is that all the chemical elements composing it would be in a state of utter dissociation, and converted back into the original protyle from which they were derived, again to be slowly reconverted into their former atomic condition as the temperature fell.