"On the other hand, when we consider the spaces in which the planets move, we find them entirely void, or as good as void; for, if there is any matter in them, it is so rare as to be without effect on the planetary motions. There is, therefore, no material connection now existing between the planets through which they might have been forced to take up a common direction of motion. How, then, are we to reconcile this common motion with the absence of all material connection? The most natural way is to suppose that there was once some such connection, which brought about the uniformity of motion which we observe; that the materials of which the planets are formed once filled the whole space between them. There was no formation in this chaos, the formation of separate bodies by the mutual gravitation of parts of the mass being a later occurrence. But, naturally, some parts of the mass would be more dense than others, and would thus gather around them the rare matter which filled the intervening spaces. The larger collections thus formed would draw the smaller ones into them, and this process would continue until a few round bodies had taken the place of the original chaotic mass." Kant, however, failed to account satisfactorily for the motion of the sun and planets. According to his system, all the bodies formed out of the original nebulous mass should have been drawn to a common centre so as to form one sun, instead of a system of revolving bodies like the solar system. 388. _Herschel's Hypothesis._--The idea of the gradual transmutation of nebulæ into stars seems to have been suggested to Herschel, not by the study of the solar system, but by that of the nebulæ themselves. Many of these bodies he believed to be immense masses of phosphorescent vapor; and he conceived that these must be gradually condensing, each around its own centre, or around the parts where it is most dense, until it should become a star, or a cluster of stars. On classifying the nebulæ, it seemed to him that he could see this process going on before his eyes. There were the large, faint, diffused nebulæ, in which the condensation had hardly begun; the smaller but brighter ones, which had become so far condensed that the central parts would soon begin to form into stars; yet others, in which stars had actually begun to form; and, finally, star-clusters in which the condensation was complete. The spectroscopic revelations of the gaseous nature of the true nebulæ tend to confirm the theory of Herschel, that these masses will all, at some time, condense into stars.