here it is also familiar as an object which may be seen at the antipodes, thus showing that it extends as a ring around the entire globe, following almost exactly a great circle of the heavens. This is the so-called milky way, which all astronomers tell us consists of tens of millions of separate stars, many of them probably equal to, and some of them larger than, our sun. It is only necessary to look in the same direction on any clear summer night, as twilight is giving place to darkness, before the small stars are visible, to see that almost all the stars then visible lie in a long broad belt from Alpha Centauri to Orion. This is so striking that if you ask anyone what is the distribution of the stars, this feature cannot fail to be observed as a stream of stars. Wait a little longer and observe the milky way: one end of the stream of stars will be found to lie on it, but makes a very small angle with it. Again the winter, aspect of the milky way suggests exactly the same idea, with this striking addition, that parts appear to start away from the main ring in a series of streams frequently corresponding with sprays of stars. Again, Proctor says, “that the stars of the first six orders are gathered into two definite regions, a northern and southern, so markedly, that the distribution of stars within these regions is richer than the distribution over the rest of the heavens in the proportion of five to two.” Thus the mere naked-eye appearance of the heavens points to its being a definite system, and the older philosophic writers have often called attention to this fact. Wright, Kant, Huygens, and many others expressed themselves strongly on the order of the heavens, and appear to have had no doubt of its being either one or more systems, and several have classified these systems into various orders, of which the visible universe does not appear necessarily to form the highest order. Kant says, speaking of the systems really known, “we trace here the first terms of a series of worlds and systems, and these first terms of an infinite series enable us to infer the nature of the rest of the series.” But if the naked-eye view gives it the appearance of a definite system, it will be seen that telescopic observations demonstrate the fact. Sir J. Herschel, who studied star-distribution more that any other man, says that the mass of stars is generally flat, of small thickness. He also says, that the number of stars visible in his telescope in the milky way number about eighteen millions, and about two millions in the remainder of the celestial vault. Struve published a list of stars in which he showed that in equal areas there were 4 1/4 at the poles of the galaxy to 122 in the galaxy itself. Herschel also says in another place—That beyond a certain magnitude all the stars lie in the milky way. There is another feature of the heavens which the telescope reveals to us, namely, the nebulæ at the poles of the galaxy, and the star-clusters in the galaxy itself. Mr. Cleveland Abbe, from Herschel's catalogues, says:—“Imagine a belt thirty

degrees wide, extending around the heavens, including the milky way. ^{* * *} This belt will include one-fourth the surface of the celestial sphere. ^{* * *} Here we find nine-tenths of the star-clusters, and one-tenth of the nebulæ.” In another paper I shall attempt to show that most likely these few nebulæ are not of the same order as the polar nebulæ.

Proctor, who discussed these facts very fully, after showing that star-clusters essentially belong to the galaxy, and, as we pass from that great circle, we go through regular stages of lessening solvability to the galactic poles, and there the nebulæ are completely irresolvable, says:—“I believe that cause may be assumed not unreasonably to be the difference in the circumstances under which the galactic and extra-galactic nebulæ have reached their present state.” Again, in respect to the nebulæ at the poles of the milky way, Proctor shows that every theory of their existence is ridiculous, “unless we concede that the nebulæ belong for the most part to our galactic system.” The accompanying charts and sketches by Sidney Waters, Proctor, and Newcombe, show strikingly this most remarkable arrangement. Thus, so far, we see that the milky way is a region of stars and star-clusters, and that the poles of this ring are regions of nebulæ. I shall now show that our sun occupies roughly the centre of this system, in a region poor in stars. Proctor says, after a very long discussion of the question, “all these phenomena point to the conclusion that the milky way, in this neighbourhood at any rate, is really what it appears to be—a belt or zone of stars, separated from us by a comparatively starless interval.” After discussing various hypotheses, he says:—“In either case we must assume that our sun is not very far from the centre of the system.” The picture of the universe we obtain from these extracts is a clear and distinct one. But perhaps the most striking argument that has yet been offered for the common origin of the universe is that of the spectroscope; which shows identity in the composition of the sun and stars with the elements to be found in the earth. The analyses of meteorites, in which no extra terrestrial element has ever been found, clearly point to the same conclusion.

I will give the opinions of a few astronomers on the evidence I have offered. Sir William Herschel distinctly states, that any sound theory of the universe must account for the peculiar arrangement of the nebulæ. Proctor, in speaking on grounds of probability, says:—“Where the results are in direct contact, the rich regions for one order corresponding to the poor regions for the others, and vice versâ, the two orders of objects belong to one system,” and again says that he knows of no single reason for supposing these nebulæ to be external galaxies. That nebulæ are not external galaxies is proved from the facts recorded by Schmidt, Hind, and others,

that nebulæ have been observed to vary and disappear, which is clearly impossible with a galaxy like our milky way. I will only give one other extract from among a large number. It is from the same work (“Proctor on the Universe”), which is wholly devoted to demonstrate these conclusions. “The phenomena I have been discussing seem to point to conclusions very different from those which have been usually accepted respecting the visible universe. Instead of separating the stars and nebulæ into distinct systems, or rather of looking on the stellar system as a member of the system of nebulæ, we seem compelled to look on almost every object visible even in the most powerful telescope, as a portion of one system, which comprises within its range, simple, multiple, and clustering stars, irresolvable nebulæ, gaseous bodies of symmetrical and unsymmetrical figure, and in all probability myriads of other forms of matter as yet undetected.”

These are the more general conclusions as to the constitution of the heavens. There is a great deal of special evidence pointing the same way; but I can only mention it here. I refer to the fact that nearly all the temporary and variable stars are in the milky way, the community of motion of groups of stars, the tendency to stream formation, and the special character of the milky way nebulæ. But what I have already discussed is sufficient to show distinctly that our universe is one system of definite construction. It undoubtedly consists of a ring or spiral of stars, star-dust and star-clusters. About the centre of this ring our sun is situated, in a comparatively sparsely-spread region. If we suppose a line to pass through our system, at right-angles to the plane of the galaxy, it passes in each direction through a region of thousands of nebulæ—these nebular masses being, as it were, polar caps covering approximately one-sixth the celestial sphere. It is certain that such an arrangement is absolutely incompatible with a chance distribution, and that consequently it offers a perfectly legitimate ground for scientific induction. In offering this hypothesis I do so with the more confidence as it is probable that every wide generalization tends to give direction to much successful research, the results of which are of great value, although some of these may convert the hypothesis into mere scaffolding, to be removed when the structure is complete. I shall assume the existence of large bodies without discussion, as such discussion is antecedent to the especial purpose of this paper, and besides, would unduly increase its length, which is too great already. I propose to discuss their claim on your consideration in a future paper. I will, however, call your attention to the present views of mathematical physicists, which point to the final state of the universe being one gigantic body, with all the energy dissipated as uniformly diffused heat. If, therefore, we may look forward to such a body in the future, why not in the past? I will not, however,

go quite so far as to suppose the body cold. I shall assume two stupendous bodies, having small independent proper motion in space, being probably at exceedingly high temperatures, endowed with a considerable rotation, and having a large number of bodies revolving around them, and not unlikely making up a considerable proportion of their mass. The probable existence of such bodies is rendered likely on the view of the cosmogony which follows as a logical deduction from the conception of partial impact.

I will now place before you some of the broader conclusions which are general deductions to all cases of partial impact:—

I have now given a picture of the universe, drawn from absolutely independent observations, and also, I believe, a set of logical deductions from “partial impact,” and it will be in the knowledge of many here that these deductions were worked out before I had studied the construction of the heavens, as exhibited by modern research. We will now compare the result of astronomical observation with the theory of partial impact.