mentioned the small mean density of the giant planets-we have at once the strongest possible evidence to show that the condition of these bodies must be unlike that of the earth. Of course, if we assume that Saturn's substance (to limit our attention to this planet) is composed of materials altogether unlike any which exist on earth, a way out of our difficulty is found, though not an easy one. In that case, however, we are only substituting one form of complete dissimilarity for another. And all the results of spectroscopic analysis, as applied to the celestial bodies, tend to show the improbability that such differences of elementary constitution exist-we will not say in the solar system only, but in the sidereal universe itself. If, however, we admit that Saturn is in the main constituted of elements such as we are familiar with, we find it extremely difficult, or rather it is absolutely impossible, to suppose that the condition of his substance is like that of the earth's. There are certain unmistakable facts to be accounted for. There is the mighty mass of Saturn, exceeding that of the earth ninety-fold. That mass is endued with gravitating energy, precisely in the same way as the earth's mass. There must be from the surface toward the centre a continually increasing pressure. This pressure is calculable,' and enormously exceeds the internal pressures existing within the earth's interior. There is no possibility of cavities, as Brewster and others have opined; for there is no known material, not the strongest known to us, iron, or platinum, or adamant, which could resist the pressures produced by Saturn's internal gravitation. Steel would be as yielding as water under these pressures. There must be compression with its consequent increase of density, such compression exceeding many million-fold the greatest with which terrestrial experimenters have dealt. That, with these enormous forces at work, the actual density of Saturn as a whole should be far less than that of water is utterly inexplicable, unless Saturn's condition be regarded as altogether unlike that of the earth. We see in the sun an orb which, notwithstanding its enormous mass, has a mean density much less than the earth's, and little greater than that of water; but we have no difficulty in understanding this circumstance, because we

1 It is a misfortune for science that Newton never published the reasoning which led him to the conclusion that the earth's mean density is equal to between five and six times the density of water. This, as every one knows, has been confirmed by several experimental methods; and, so far as appears, the problem is a purely experimental one. Newton, however, made no experiments; at least, none have been heard of as effected by him, and it is scarcely probable that he had any instruments of sufficient delicacy for a task so difficult. Prof. Grant ascribes Newton's conclusion to a happy intuition; yet it is very unlike Newton to make a guess on such a matter. It is more probable that he guessed the elements of the problem than the result. He probably assumed that the earth's mass is composed of a substance like granite, and, adopting some law of compression for such a substance (based on experiment, perhaps), calculated thence the compression at different depths, and so obtained the mean destiny of the whole mass.

see that the sun is in a state of intense heat, and we know that this heat produces effects antagonistic, as it were, to those produced by the attraction of his mass as a whole upon every portion of his substance. But, if we make no similar assumption in Saturn's case, we find his small density inexplicable.

Another circumstance associated with the question of Saturn's density introduces new difficulties of the most perplexing nature if it be regarded according to the ordinary view, while it seems not only explicable, but manifestly to be expected, on the theory that Saturn's whole orb is in an intensely heated condition. Saturn certainly has an atmosphere of considerable depth. The belts which surround his globe are evidently produced by clouds in his atmosphere, though what the nature of these clouds may be is not as yet known. The brighter belts are the cloud-belts, while the darker either show his real surface, or, far more probably, belong simply to lower cloudlayers. These belts are variable in appearance and position, sometimes changing with great rapidity. Their real extent is enormous, exceeding the whole surface of our earth, even in the case of the narrowest belts yet seen. No one who has viewed them through telescopes of great power can refuse to adopt the conclusion that the atmosphere in which these great cloud-zones are suspended must be of great depth, certainly far deeper than our atmosphere. But such an atmosphere, subjected to the attractions of Saturn's mass, would be enormously compressed underneath those manifestly thick cloud-layers. A very moderate assumption as to the depth of the atmosphere would lead to the conclusion that at its base it must be denser than water-that is, denser than Saturn himself. No gas could exist as gas at this density. Apart from this, we are here arriving at the very theory which the ordinary view of Saturn teaches us to avoidviz., the theory that he is utterly unlike our earth in physical condition. We may much more conveniently arrive at the same general conclusion, while avoiding other difficulties, by simply adopting the same explanation in this case which serves to account also for the small density of Saturn's mass-viz., the theory that Saturn's globe is in a state of intense heat.

But now let it be noticed how perfectly this view of Saturn's con dition accords with the theories which are beginning to be established respecting the genesis of the solar system. Whether we regard the planets as formed from the condensation of enormous nebulous masses, or whether we assume that they were produced by the gathering together of matter originally traveling in dense meteoric flights around the central aggregation whence the sun was one day to be formed, we see that the larger the planet the greater must have been its origiual heat. The heat generated during the condensation of a nebulous mass must depend upon the magnitude of the mass, since in fact the accepted theory of heat teaches us that the original heat of a globe so

formed is measurable by the actual difference in dimensions between the globe and its parent cloud-mass, and of course the larger the cloud-mass the greater this difference would necessarily be. It is equally certain that the heat generated by the gathering-in of meteoric matter would be so much the greater according as the quantity of matter gathered and gathering was greater; for the heat is produced by the downfall of such matter on the globe it helps to form, and the greater the mass of that globe the greater is its attracting might, the greater the velocity it generates in the falling meteors, and therefore the greater the heat produced when they are brought

to rest.

Saturn, then, would originally be much hotter than our earth, on any theory of the evolution of our solar system-and there are few astronomers who doubt that the solar system was wrought by processes of evolution to its present condition. But not only would Saturn be much hotter than the earth, but, owing to his enormous size, he would part with his heat at a much slower rate. On both accounts we should infer that at this present time Saturn is much hotter than the earth— in other words, since our earth still retains no inconsiderable proportion of its original heat, Saturn may be assumed to be in a state of intense heat. What his actual heat may be is not so easily deter mined. We shall presently show reasons for believing that an inferior limit, below which his heat does not lie, is indicated by the fact that he still possesses inherent luminosity. On the other hand, a superior limit is indicated by the fact that his inherent luminosity is not great, and that, in all probability, the thicker cloud-zones of Saturn prevent the passage of the greater part of his light.'

We should infer, then, that Saturn in some respects resembles the sun, though of course the very same reasoning which teaches us to believe that Saturn is very much hotter than the earth, leads us also to the conclusion that it is not nearly so hot as 'the sun. Now, thus viewing Saturn, we should be led to expect, apart from all telescopic evidence to that effect, that he would resemble the sun in certain general features. For instance, we might expect that he would have spotzones, while his equatorial zone would be free from spots; or, if it were thought that so close a resemblance was not to be looked for, then we might still expect that his equatorial zone, like the sun's, would be distinguished from the rest of his surface by some wellmarked peculiarity. This is the case. The equatorial zone of Saturn is distinguished by a peculiar brightness from the rest of his surface, insomuch that the late Prof. Nichol was led to regard this zone as the

1 To prevent misapprehension, it may be as well to remind the reader that the apparent continuity of Saturn's cloud-belts by no means implies that they are really continuous, and it is on a priori grounds highly improbable that they are so; openings in his cloud-zones two or three hundred miles in length and breadth would be quite undiscernible at Saturn's enormous distance.

scene of a constant precipitation of meteoric matter from the inside of the ring-system.

Now, there is one important peculiarity which distinguishes the equatorial bright zone of Saturn from that of Jupiter. Jupiter's axis is almost square to the level of the path in which he travels around the sun; so that his equatorial zone lies nearly in that level, and is therefore directly illuminated by the sun. The aspect of Jupiter in fact, as seen from the sun, is always that which our earth presents in spring and autumn. But Saturn has an axis very considerably sloped to the level of the path in which he travels. It is more sloped even than our earth's axis. So that in the course of his long year of 10,759 days (29) of our years) Saturn's globe presents toward the sun all the varying aspects which our earth presents, only with a somewhat greater range of variation. At one time he is placed as our earth is in spring, and then his equatorial belt, as seen from the sun, appears to lie straight across the middle of his disk. Rather more than seven years later he is posed as our earth is posed at midsummer, his northern pole is bowed toward the sun, and his equator is seen as a half-oval, curving far south of the middle point of his disk. He passes on from this position, and in seven more years he is placed as our earth is in autumn, with his equator again lying straight across his disk. Then, seven years or so later, he presents the aspect of our earth at midwinter, his equator curved into a half-oval passing far to the north of the middle point of his disk. And, finally, at the end of yet seven years more (or, more exactly, of one complete Saturnian year from the commencement of these changes), he is again as at first. Now, it seems manifest that, if the great cloud-zone which surrounds Saturn, appearing as a nearly white ring, were due to solar action, it would fluctuate in position as these changes proceeded. The very length of the Saturnian year should insure the occurrence of such fluctuations. We have only to inquire what takes place on our own earth, where, though we have nothing comparable with the belt systems of Jupiter and Saturn, we have, nevertheless, over ocean-regions, a sun-raised tropical cloud-band in the middle of the day. This cloud-band follows the sun, being equatorial in spring, passing far north of the equator, even to the very limit of the torrid zone, in summer, returning to the equator in autumn, passing to the southern limit of the torrid zone in winter, and returning again to the equator in spring. In fact, this cloud-band as seen from the sun would always cross the middle of the earth's face as a straight line in spring and autumn, and as considerably more than a half-oval, agreeing in position with the tropics of Cancer and Capricorn, at midsummer and midwinter. But nothing of the sort happens in Saturn's case. His equatorial white ring is really equatorial at all times, instead of being drawn to his tropics at his midsummer and midwinter seasons. This, in our opinion, is decisive of the origin of this great band. If it were sun-raised, it would obey the sun; but, being raised

by Saturnian action, its position is solely determined by Saturn's rotation, and it therefore remains constantly equatorial.

But next a very strange and, at a first view, incredible circumstance has to be considered in immediate connection with the relations we have been dealing with.

It sounds startling to suggest that Saturn probably changes at times in size and shape. Yet the evidence in favor of the suggestion is very weighty. It may briefly be presented as follows:

In April, 1805, Sir William Herschel, who had hitherto always seen the planet of an oval figure, found that it presented a strangely distorted appearance. It was flattened as usual at the poles, but also at the equator; accordingly, it had a quadrangular or oblong figure (with rounded corners, of course), its longest diameters being the two which (crossing each other in the middle of the disk) passed from north latitude 43° on Saturn to the same southerly latitude. Or we may otherwise describe the appearances presented, by saying that Saturn seemed swollen in both the temperate zones. Herschel found that the same appearance was presented, no matter what telescope he employed, and he tried many, some seven feet, some ten, one twenty, and one forty feet in length. With these telescopes Jupiter presented his ordinary oval aspect. But Herschel is not the only astronomer by whom such appearances have been noticed. On August 5, 1805, Schröter found that Saturn's figure was distorted. Dr. Kitchener says that in the autumn of 1818 he found Saturn to have the figure described by Herschel. The present Astronomer Royal has seen Saturn similarly distorted, and on another occasion flattened in the temperate zones. In January, 1855, Coolidge, with the splendid refractor of the Cambridge (U. S.) Observatory noticed a swollen appearance in Saturnian latitude 20°; yet on the 9th the planet had resumed its usual aspect. In the report of the Greenwich Observatory for 1860-'61, it is stated that "Saturn has sometimes appeared to exhibit the square-shouldered aspect." The two Bonds, of America, surpassed by few in observing skill, have seen Saturn square-shouldered and have noticed variations of shape. It seems impossible to reject such testimony as this. Nor can it be disposed of by showing that ordinarily Saturn presents a perfectly elliptical figure. It is the essential point of the circumstances we are considering, that they are unusual.

Now, we do not pretend to explain how such changes of shape are brought about. But we would invite special attention to the circumstance that if these changes be admitted as having occasionally occurred (and we do not see how they can be called in question), then the result is only startling in connection with that theory of Saturn's condition which we are here opposing. If Saturn be a globe resembling our earth, then sinkings and upheavals, such as these appearances indicate, must be regarded as involving amazing and most stupendous throes-as in fact absolutely incredible, no matter what evidence may

VOL. IV.-4