with all or even the generality of comets it would be assuming too much to assert.

The most remarkable feature of modern astronomical discovery remains yet to be mentioned. A phenomenon which men had long been in the habit of looking upon as a meteorological one has been at length recognized in its true light, and has been found wonderfully to enhance our appreciation of the complexity of the systems which exist within the solar domain. Meteors, shootingstars, and aerolites have taken their place among the attendants of the Sun; and, in several instances, the orbits they have followed before they reached the Earth have been approximately determined.

But it is rather as members of systems than as individual bodies, that these objects acquire their chief interest and meaning. There was not much, perhaps, to attract attention to them when they were supposed to form one or two rings occupying a position in space very nearly coincident with that of the earth's orbit. But it has now been placed beyond a doubt that the earth encounters fiftysix systems, at least, of these small bodies. And these systems are found (in the only instances yet examined) to be-not circular rings-but ovals of great eccentricity extending far into space, even, in some cases beyond the orbits of Uranus and Neptune. It is clear, then, that we can no longer look on these systems as resembling, in the remotest degree, the asteroidal zone. We are forced, too, to take into consideration an important question of probability. What is the likelihood that if there were but a few hundreds of such systems, the earth would encounter so many as fifty-six? The probability may be reckoned "almost at naked nothing." And therefore we are compelled to admit as a legitimate à posteriori deduction, the extreme probability, we may almost say the certainty, that such systems are to be reckoned-not by hundreds and thousands-but by millions on millions.

Nor is this all. Within the last few months the startling discovery has been made that two of the meteoric systems at least, and probably many others, coincide throughout their calculated extent with the orbits of known comets. Accordingly, we are led to trace an intimate connection, if not an absolute identity, between comets and shooting-star systems. And when we find that a system, which has afforded such grand displays of star-falls as the well-known November shooting-star system, is identified-not with a large and conspicuous comet-but with one which has only lately been detected, though it must have been in close proximity to the earth some thirty times during the last thousand years, with a comet,

in fact, which is absolutely invisible to the naked eye, and far from being a conspicuous object in powerful telescopes, we are led to recognize the importance of such comets as Newton's, Halley's, and Donati's.

The result to which these considerations lead is clearly this :

The interplanetary spaces, so far from being looked upon as untenanted, save by an occasional wandering comet, must be considered as crowded with various forms of cosmical matter. I would not be understood as using the term "crowded" in a sense implying absolute proximity between the various members of the cometic or meteoric systems. On the contrary, the evidence that we have assures us that the sum of the volumes of all the members of a system must bear an indefinitely small proportion to the total space occupied by the system. But if an eye armed with new powers of vision, and placed at some far distant point, could see at one glance all the systems which occupy the solar domain, they would appear as a complicated network formed by interlacing streams of cosmical dust. And, amidst the streams of misty light representing cometic or meteoric systems, the planets would shine forth as distinctly and as brilliantly as the brighter stars upon the background of the Milky Way.

Nor does it seem difficult to determine the general laws according to which the density or compactness of the interwoven streams would seem to an eye placed as we have supposed—to vary with distance from the central orb.

Proceeding from the outermost parts of the system towards the Sun, we may conceive that there is in the more remote regions a gradual condensation, but that this increase of density becomes much more rapid in the immediate neighbourhood of the Sun. This is not the only law, however, according to which the density of matter distributed throughout the solar domain must be supposed to increase. There is a certain plane near which all the primary members of the solar system are observed to move. In the asteroidal family, whose members depart more freely than the larger planets from this great central plane, there is, nevertheless, a distinctly marked obedience to the general law of aggregation in its neighbourhood. Three-fourths of the asteroids revolve in orbits less than ten degrees inclined to the medial plane, and there are not ten of them whose orbits are inclined so much as twenty degrees. Now, on a first view of cometic orbits, we notice scarcely any trace of a tendency to aggregation near the medial plane of the solar system. Nay, so far is this from being the case, that

among several hundreds of comets whose orbits have been determined, a decided tendency of a very different character has been noticed. If we suppose the Sun to lie at the common vertex of a double cone having a semi-vertical angle of forty-five degrees, and its axis perpendicular to the medial plane, then the planes of cometic orbits exhibit a tendency to present themselves as tangent-planes to this imaginary cone."* But when we consider cometic orbits more closely, we find abundant evidence of a tendency amongst those comets which are nearest to the Sun to aggregate around the medial plane of the solar system. There are some twenty comets which have been recognized as travelling within the orbit of Saturn. Among these there are only two whose orbits are inclined more than fifteen degrees to the medial plane of the solar system. Now there is no reason whatever for supposing that there are not multitudes of undetected comets whose perihelia lie far nearer to the Sun than any yet discovered. On the contrary, we have distinct evidence of a rapid increase in the number of perihelia, with decrease of distance down to and within the neighbourhood of the Earth's orbit ;† and, remembering the probability that comets whose perihelia lie nearer to the Sun would escape observation altogether, we have every reason for supposing that this law of increase is continuedas why should it change?-right up to the immediate neighbourhood of the Sun. And further, we may confidently assume that that obedience to planetary laws which, as we have seen, begins to be

This tendency has been pointed out by one of our most distinguished modern astronomers. As it seems impossible to suggest any rational explanation of so remarkable a peculiarity-for we have to explain, not merely the fact that the orbit-planes show no tendency to coincidence with the medial plane, but also the fact that the medial plane should be connected, in any way, with cometic orbits-one seems permitted to question whether the peculiarity is real, or only apparent. Now, if we remember that, cæteris paribus, the greater the inclination of a comet to the plane of the ecliptic (virtually coincident with the medial plane of the solar system) the greater the antecedent probability that the comet will be detected, we may recognize a cause for the observed peculiarity, independently of any real peculiarity in the arrangement of cometic orbits. A gradual diminution in the number of orbits as we leave the plane of the ecliptic, in combination with this gradual increase in the probability of detection, might very well lead to such a result as we have specified. According to the law of decrease or increase respectively, the variation in the numerical distribution of observed cometic orbits might point to the existence of a maximum at any assigned inclination to the ecliptic.

The tendency in question is so far from being strongly marked that this consideration may, for the present, be held to be a sufficient explanation.

Thus, out of one hundred observed comets, fifty-eight have perihelia between forty millions and a hundred millions of miles from the Sun, twenty have a less perihelion distance, and four only have a perihelion distance exceeding one hundred and sixty millions of miles.

exhibited by comets within the orbit of Saturn, becomes yet more marked among comets nearer to the Sun. Therefore, it seems highly probable that cometic orbits, and especially those which are nearest to the Sun, show a marked tendency towards aggregation near the medial plane of the solar system.

A celestial phenomenon, of which we have not hitherto spoken, appears to gain a far easier explanation from the considerations above adduced, than from the theories ordinarily adopted respecting it. The zodiacal light has been accounted for in three ways. There are some who hold that it is an atmosphere of the sun; others that it consists of a ring of cosmical particles, travelling around him in a nearly circular orbit; and others that it consists of a lenticular disc of cosmical dust, each portion of which travels in a nearly circular orbit. The remarkable phenomena presented by the zodiacal light, its strangely fluctuating figure, its varying position, and the singular increase and diminution noticed in its distinctness, are not accounted for by any of these theories. But if we recognized in the zodiacal light merely the effect of the above-considered aggregation among the cometic or meteoric systems which exist within the solar domain, the variations I have mentioned become readily explicable. A multitude of bodies travelling in orbits of every degree of ellipticity and magnitude, but with a marked aggregation in the neighbourhood of the Sun, and with a yet more marked aggregation in the neighbourhood of the medial plane of the solar system, would, in the first place, exhibit precisely such an appearance as the zodiacal light; and, in the second place, the general illumination resulting from the congregated comets would be liable to continual variation. Comets would be continually arriving within and passing away from the region within which their light would assist in forming the appearance we are considering. At one time the press of arrivals would temporarily increase the density of cometic aggregation; at another, the reverse would hold for awhile, and the zodiacal light would wax and wane accordingly, precisely as it is observed to do. So also its figure and apparent position would be liable to changes corresponding to those which are actually presented. Therefore, without denying positively that the zodiacal light is caused by the existence of a multitude of minute bodies travelling in orbits of small eccentricity around the Sun, we hold that the phenomena correspond far more closely with those which would be presented if there is in the neighbourhood of the Sun a great increase in the density with which cometic and meteoric systems are congregated together in

the neighbourhood of the medial plane of the solar system. And this correspondence becomes a strong argument in favour of such an increase of density when it is remembered that, as we have seen, there exist independent reasons for believing an aggregation of this sort to be not only possible, but highly probable.

But whatever opinion we may form on this and kindred questions, there is no dubiety whatever about the general results which have been presented above. Our conceptions of the solar domain are different, indeed, from those formed of old. "There was true

prophecy," as has been well remarked by the late Professor Nichol, "in the exclamation of Laplace, who, although knowing more of the celestial mechanism than any man then living, said earnestly, on his death bed, That which we know is little; that which we know not is immense.'"

When we turn to examine the views which were held respecting the sidereal system at the commencement of the present century we find that they are distinguished by no very marked points of difference from those at present entertained. Yet have many important discoveries been made in the interval, which seem to suggest a modification, in many respects, of the views which have so long held their ground. What these are we proceed to point out.

So soon as the Copernican theory had become thoroughly established, and had been supplemented by adequate conceptions of the dimensions of the Earth's orbit round the Sun, it became manifest that the stars must be placed at an enormous distance from the solar system. That the motion of the Earth in an orbit one hundred and eighty millions of miles in diameter, should produce no appreciable effect on the configuration of the constellations, could be explained in no other way than by supposing that an orbit of these dimensions, viewed from the nearest fixed star, would scarcely present appreciable proportions. And when the nicest observation with the most accurate instruments which were then procurable, showed that any parallactic displacements which might exist among the stars, were insensible, or, at any rate, fell short of recognised instrumental errors, astronomers were compelled yet further to extend their conceptions of the immensity of the interval which separates the Sun from the nearest fixed star.

Accordingly, the notion that each visible star may be a Sun, in magnitude and splendour equal to, or perhaps excelling our own, was early recognized by astronomers as not merely reconcilable with the apparent minuteness of the stars, but as suggested by a comparison of the brilliancy of their light at the enormous distance