It is recognised by mathematicians that within any great system of bodies subject to the law of gravitation there can be no such thing as motion of any of them in a straight line; neither can any amount of motion arise within such a system through the action of gravitation alone capable of carrying any of its masses out of the system. The ultimate tendency must be towards concentration rather than towards dispersal. It seems, therefore, only reasonable to consider whatever motions and whatever velocities we find among the stars, as having been produced by the gravitative power of the larger aggregations, modified perhaps by electrical repulsive forces, by collisions, and by the results of those collisions; and we may look to the changes now visibly going on in some of the nebulæ and clusters as indications of the forces that have probably brought about the actual condition of the whole stellar universe. If we examine the beautiful photographs of nebulæ by Dr. Roberts and other observers, we find that they are of many forms. Some are extremely irregular and almost like patches of cirrus clouds, but a large number are either distinctly spiral in form, or show indications of becoming spiral, and this has been found to be the case even with some of the large irregular nebula. Then again we have numerous ring-formed nebulæ, usually with a star involved in dense nebulosity in the centre, separated by a dark space of various widths from the outer ring. All these kinds of nebulæ have stars involved in them, and apparently forming part of their structure, while others which do not differ in appearance from ordinary stars are believed by Dr. Roberts to lie between us and the nebula. In the case of many of the spiral nebulæ, stars are often strung along the coils of the spiral, while other curved lines of stars are seen just outside the nebula, so that it is impossible to avoid the conclusion that both are really connected with it, the outer lines of stars indicating a former greater extension of the nebula whose material has been used up in the growth of these stars. Some of these spiral nebulæ show beautifully regular convolutions, and these usually have a large central star like mass, as in M. 100 Comæ and I. 84 Comæ, in Vol. II. Pl. 14 of Dr. Roberts's photographs. The straight white streaks across the nebula of the Pleiades and some others are believed by Dr. Roberts to be indications of spiral nebulæ seen edgewise. In other cases, clusters of stars are more or less nebulous, and the arrangement of the stars seems to indicate their development from a spiral nebula. It is to be noted that many of the objects classed as planetary nebulæ by Sir John Herschel are shown by the best photographs to be really of the ring-type, though often with a very narrow division between the ring and the central mass. This form may therefore be of frequent occurrence.