Having begun by assuming that the stars are moving apparently in every direction with great velocities, they proceed to point out that sooner or later, although the lapse of time may be extraordinarily long, collisions or near approaches between stars are bound to occur. In the case of collisions the chances are against the bodies striking together centrally, it being very much more likely that they will hit each other rather towards the side. The nebulous mass formed as a result of the disintegration of the bodies through their furious impact would thus come into being with a spinning movement, and a spiral would ensue. Again, the stars may not actually collide, but merely approach near to each other. If very close, the interaction of gravitation will give rise to intense strains, or tides, which will entirely disintegrate the bodies, and a spiral nebula will similarly result. As happens upon our earth, two such tides would rise opposite to each other; and, consequently, it is a noticeable fact that spiral nebulæ have almost invariably two opposite branches (see Plate XXII., p 314). Even if not so close, the gravitational strains set up would produce tremendous eruptions of matter; and in this case, a spiral movement would also be generated. On such an assumption the various bodies of the solar system may be regarded as having been ejected from parent masses. The acceptance of the Planetesimal Hypothesis in the place of the Hypothesis of Laplace will not, as we have seen, by any means do away with the probability that our solar system, and similar systems, have originated from a nebulous mass. On the contrary it puts that idea on a firmer footing than before. The spiral nebulæ which we see in the heavens are on a vast scale, and may represent the formation of stellar systems and globular clusters. Our solar system may have arisen from a small spiral.