We have also shown, at page 119, that the nebula must have been somewhat in the form of a cylinder terminated at each end by what may be looked upon as a segment of a sphere, although it would more probably be an almost shapeless mass of cosmic matter, because the greater part of it would be very slowly brought under the influence of centrifugal force as it fell in from the polar directions; and again, a few pages back, that almost all the matter coming in from its equatorial regions--even what might be called its tropical regions--would be intercepted before it could reach the Saturnian nebula. Likewise, at page 137, when examining Bode's law reversed, we have seen a limit set to the acceleration of the movement of revolution in the planets of the system as they approached the centre, because any acceleration beyond a certain limit, clearly marked out, would of necessity be within the nebula itself, and the rate of revolution would be less than that of the sun on its axis at the present day. This may be used as an argument against the nebular hypothesis, but we think we have shown in the same Chapter VII. that this is not the case. But we have still to try to account for the repeated rises and falls in density in the planets from Neptune to Mercury, or even farther; which operation causes us to bring forward, first of all, a new idea as to what the form of the nebula would come to be. [Illustration: FIG. 2.] The accompanying rough sketch (Fig. 2), drawn to a scale of one-quarter inch to 1,000,000,000 miles shows that, supposing the Saturnian nebula to have been a perfect sphere, and to have abandoned matter till the velocity of rotation came to be equal in a region corresponding to the tropical region of the earth, the cylindrical part of it would present a straight side of more than 1,000,000,000 miles in length; provided always that the general diameter of the nebula did not decrease through condensation and contraction during the operation; but as this could not be the case the length of the cylindrical part would be considerably less than that. How much less we have no means of calculating. On the other hand we have seen, when discussing, in the case of Jupiter, how matter must have been abandoned by any nebula, that from the time the original nebula began to abandon matter through centrifugal force, it must have gone on acquiring a constantly increasing length of straight side as it contracted. Thus the Saturnian nebula would begin work with the accumulated cylindrical length it had inherited from Neptune and Uranus, so that the straight side may have been very much longer than that shown by the sketch; a simple look at it is enough to make one believe that this would be the case. But this idea naturally leads us to another digression.