the innermost still very near to its parent orb; and we find Jupiter himself in a condition, judging by his mean density, corresponding to what we may suppose to have been the earth's condition when the moon, after being fully fashioned, had receded to a corresponding distance (that is, to something like the same relative distance from the earth). Mr. George Darwin has, indeed, shown in a very interesting discussion of Jupiter's compression and the motions of his moons, that from something more than the mean density of the planet this condition of Jupiter may be inferred. It is as nearly demonstrated as such a relation can well be, that the central part of Jupiter is greatly compressed, compared with the outer parts of the planet as we see him. Now Saturn, judged by his mean density, appears to be in an earlier stage of his career as a planet than Jupiter. While Jupiter's mean density is but a fourth of the earth's, Saturn's is barely one-seventh of hers. We may somewhat safely infer, then, that Saturn's moon-generating work is not so far advanced as Jupiter's; and, in fact, considering his moons only, we see that this really is so, for his nearer moons are much closer to him, absolutely as well as relatively, than are those of Jupiter. It is true Japetus, the outermost Saturnian moon, has receded to a greater absolute distance than the outermost moon of Jupiter, and Saturn being a smaller planet than Jupiter, this greater absolute distance of Japetus implies a relative distance greater in still higher degree. But this only serves to show that, from whatever cause, the moon-generating process in the case of the Saturnian system has progressed more slowly than in the case of the Jovian system. It is at any rate clear that the innermost of Saturn's moons is relatively much younger than the innermost of Jupiter's. This being so, what opinion are we to form of the ring-system? Does it not, on the face of matters, appear as though this ring-system represented an embryo moon, or perhaps the embryos of several moons? Finding thus around the planet of least density (presumably, therefore, the one which has advanced least towards its final condition), the

planet which has the nearest moons, and, in fine, the planet which-if such planet there is-must be regarded as alone in the moon-generating stage of planetary existence, this singular appendage, absolutely unique in the solar system, are we not justified in saying that here we see the last stages of the moon-producing stage of a planet's life? It seems to me that this is the most probable interpretation of the ringsif it be not the only interpretation available. If we accept it, we see what a moon is like when as yet not fully fashioned. It consists not of a single globe, not of several large globes one day to condense into one, but of rings of multitudinous bodies, strewn so closely that, from a distant observingstation, they appear to form continuous solid or liquid rings. If this were an interpretation of the Saturnian rings to which the discussion of our own moon had led us, a certain degree of hesitation might be suggested by the circumstance that possibly our interpretation so deduced might be a little forced. But the reverse of this holds-we are encouraged to adopt the view instead of being led to doubt it—when we note that more than seventeen years since, the Saturnian rings were proved to be constituted in the manner here described. Nothing can be much more complete than the demonstration of this which was given by the Bonds and Prof. B. Peirce in America, and Prof. Clerk Maxwell in England. There cannot now be a shadow of doubt that the entire Saturnian ring-system consists of discrete satellites, as the sands of the sea-shore for multitude, richly aggregated in some parts of the system's breadth, sparsely strewn in others. Now, this being presumably an embryonic moon-system-perhaps to form one moon, perhaps to form several—we have strong evidence in favour of the belief that a moon is thrown off from the parent planet in this form, and not as a single body; for certainly there is no reason for supposing that the process of moon-formation in Saturn's case would be different from the corresponding process in the case of any other planet.

We may draw yet another inference from the giant

planets as to the past of the earth and moon. So far as we can judge from Jupiter and Saturn, a planet remains in a partly vaporous, partly fluid state long after the moons have been formed, and have receded to a great distance from the parent orb. We may perhaps assume, indeed, that it is while a planet is in such a state that the forces thrusting a moon away from its parent planet's neighbourhood are most active. Unquestionably, when a large part of a planet is fluid, so that a great wave of fluid or plastic matter circuits around the planet, while what is one day to become the planet's ocean exists only in the form of steam or cloud or falling rain-showers, the forces at work checking the rotation of the planet, and pari passu repelling the infant moon, would be far more active than when the chief retarding agent was an oceanic tidal wave. Combining with this the consideration that when the moon was nearer its tide-raising power was greater-not as the inverse square, but as the inverse cube of the distance-we must attribute to the earlier stages of a moon's independent career the greater part of its work in checking the rotation of its parent planet, and thus (indirectly) causing its own repulsion from that body's neighbourhood.

Thus I cannot for my own part consider that much of the work done by the tidal wave in forming the earth's crust was effected, as Dr. Ball believes, when the moon was much nearer to the earth than now. That within the range of time over which the geologic record extends the moon's action was much more effective than it is at present, we may well believe; but that, at any time while the earth's fossiliferous strata were being formed, the moon was within 40,000 or even 100,000 miles from the earth I cannot regard as likely, or even credible. If the long time-intervals necessary to explain the features of the earth's crust could be greatly shortened by such considerations as Dr. Ball has eloquently urged, the case would, perhaps, be different. There is an enormous difficulty, unquestionably, in reconciling the vast period (Ico millions of years at least) during which the earth

seems to have been acted upon by the solar rays as at present, with the comparatively short period (not more than twenty millions of years) during which the sun can have done such work as at present, if his emission of heat is regarded as solely due to his contraction to his present dimensions. But we cannot evade the difficulty by appealing to the moon's former tide-raising energies. There are other lines of argument besides Dr. Croll's by which the vastness of the period during which the sun has worked as he does now in the emission of heat and light can be demonstrated. Either our interpretation of the source of his heat is incorrect (or at least incomplete), or else, as for my own part I believe, the process of solar contraction has gone much farther than those infer who imagine that the sun's real globe is nearly of the dimensions of that orb which is bounded by his photosphere or light-surface. But be this as it may, there can be very little doubt that when the moon was but 40,000 miles from the earth's surface, there was no life on the earth, no surface which could support life. In all probability she was in the same state as Jupiter-her surface so hot that the waters which were one day to form her oceans were kept constantly by intensity of heat in the form of vapour, save where, at a great height from the fiery surface below, they were condensed to the form of visible clouds.

As regards the future of the moon, in which is involved to some degree the future of the earth, we may accept the general conclusions of Mr. Darwin and Dr. Ball, though the estimate of the time-intervals which must elapse, ere the successive changes are reached, cannot be regarded as trustworthy. (The problems involved are far too complex to be satisfactorily dealt with in the present stage of the discussion: I doubt even whether science will have ascertained, a thousand years hence, the true rate at which the moon's recession will take place during the next ten millions of years.)

In the first place it is to be noted that the terrestrial day is now shortening more quickly (we ought rather perhaps to

E

say less slowly) than the lunar month is lengthening-so that, though the month is lengthening, the number of days it contains is gradually diminishing. It was otherwise in the past. The number of days in the lunar month continually increased until the time when the month lasted about twenty-nine days, since which time the number of days in the month has continually diminished. Dr. Ball describes the time when the number of days in a lunar month was at its maximum as the time when the month was in the zenith of its glory,—why, this deponent sayeth not, not knowing. Measured in any other way than by terrestrial days, the month grows constantly longer, and will do so until the moon no longer has any work to do in retarding the earth's rotation. This is the same as saying that the lunar month will continue to lengthen as long as it differs from the terrestrial day. Thus, great as the period would be during which the day would have to lengthen to equal the present lunar month, we have to look forward to a still greater distance in the remote future for the time when the lengthened day, and the less lengthened lunar month, will be equal. At that time the day and the month will each last 1,400 hours, as hours are now measured, or 583 of our present days. Dr. Ball puts the time when this change will have been effected 150,000,000 years from the present time. It appears to me his estimate falls far short of the truth. The actual lengthening of the day, noted since the time of Hipparchus, has accrued at a much slower rate than Dr. Ball's estimate would imply. However, even if the epoch be no more remote than this, we need not fear that the progress of the change will seriously affect either ourselves or our descendants for many generations to come. Probably long before ten millions of years have elapsed, much more important changes will have affected the earth-as loss of solar heat, the effect of long-continued internal changes, such as are now in progress, and so forth. We may be as easy respecting the lengthening of the terrestrial day, on account of the great remoteness of the final condition, as