But, in the latter part of his life, Sir William Herschel realised that this was not the true explanation of the features presented by the Galaxy. The brilliant spots and patches in it, the dark rifts and openings, the narrow streams of light often bounded by equally narrow streams or rifts of darkness, render it quite impossible to conceive that this complex luminous ring has the form of a compressed disc extending in the direction in which we see it to a distance many times greater than its thickness. In one very luminous cluster Herschel thought that his telescope had penetrated to regions twenty times as far off as the more brilliant stars forming the nearer portions of the same object. Now, in the case of the Magellanic clouds, which are two roundish nebular patches of large size some distance from the Milky Way in the Southern Hemisphere and looking like detached portions of it, Sir John Herschel himself has shown that any such interpretation of its form is impossible; because it requires us to suppose that in both these cases we see, not rounded masses of a roughly globular shape, but immensely long cones or cylinders, placed in such a direction that we see only the ends of them. He remarks that one such object so situated would be an extraordinary coincidence, but that there should be two or many such is altogether out of the question. But in the Milky Way there are hundreds or even thousands of such spots or masses of exceptional brilliancy or exceptional darkness; and, if the form of the Galaxy is that of a disc many times broader than thick, and which we see edgewise, then every one of these patches and clusters, and all the narrow winding streams of bright light or intense blackness, must be really excessively long cylinders, or tunnels, or deep curving laminæ, or narrow fissures. And every one of these, which are to be found in every part of this vast circle of luminosity, must be so arranged as to be exactly turned towards our sun. The weight of this argument, which has been most forcibly and clearly set forth by the late Mr. R.A. Proctor, in his very instructive volume _Our Place among Infinities_, is now generally admitted by astronomers, and the natural conclusion is that the form of the Milky Way is that of a vast irregular ring, of which the section at any part is, roughly speaking, circular; while the many narrow rifts or lanes or openings where we seem to be able to see completely through it to the darkness of outer space beyond, render it probable that in those directions its thickness is less instead of greater than its apparent width, that is, that we see the broader side rather than the narrow edge of it.