been led to regard meteor systems, circuiting close around the sun, as affording the true explanation of the solar spots. The late Professor Benj. Peirce, in America, and earlier, our great astronomer Sir John Herschel, propounded the idea that spots are caused by meteors falling upon the sun. "According to this view," says Professor Young (we prefer to quote the authority of others respecting a theory which we have already abundantly advocated in our own words), "the periodicity of the spots would be simply accounted for by supposing the meteors to move in a very elongated orbit, with a period of 11'1 years, adding the additional hypothesis that at one part of the orbit they form a flock of great density, while elsewhere they are sparsely distributed."

Now, here the question arises what sort of approach the meteoric orbit should make to the sun's surface,—if that can be called a surface which constitutes the visible globe of the sun. Do the meteors of this supposed system pass simply very near to the sun's surface, the outskirting bodies only being captured at each return of the main flight? Or do they actually rush through the surface which astronomers call the solar photosphere? In one case we find it difficult to understand how so great a disturbance as the period of great sunspots indicates can be produced by bodies so small and presumably so few; in the other case the difficulty is to understand how there can be more than one circuit of the meteoric stream, seeing that the passage of the whole flight through the sun's actual vaporous substance within the photosphere, should end, one would imagine, in the destruction of every meteor in the system.

Then another difficulty arises. The sun-spots, as all by this time know, appear along two zones of the sun's surface corresponding to the temperate and subtropical zones on the surface of the earth. None are ever seen on the sun's equator; none are ever seen at the solar poles. The zones along which spots may be seen, in greater or less frequency, have a tolerably wide range upon the sun's surface; and spots appear at the time of maximum sun-spot frequency, over nearly the whole of these zones, and on the northern as well as on the southern solar hemisphere.

Now, if we imagine a meteoric mass travelling on some orbit or other whose point of nearest approach to the sun lies close to the sun's surface, it is manifest that, supposing the sun-spot disturbance produced by that mass takes place when the meteor is at that point of nearest approach, the spot must appear in a certain definite place, which will be the same for every meteor travelling in the same path. Suppose, for instance, the point of nearest approach of the

meteoric orbit to the sun were in solar latitude 30° north, a sun-spot caused by any meteor travelling along that orbit would be in that solar latitude and no other. If the meteor orbit just touched the sun's surface, the same would hold. If the meteor orbit intersected the sun's surface, the point where a meteor went in would be in a different latitude from the point where the meteor would come out— if it came out. But how could it possibly come out? It would reach the surface in the form of vapour, and in that form would be most thoroughly absorbed by the sun, assuredly never finding its way again to the surface, except in its future movements as part and parcel of the sun's vaporous and gaseous globe.

So that a single meteor stream could not under any conceivable conditions account for the occurrence of sun-spots at the time of maximum disturbance in both solar hemispheres, and over wide zones of the sun's surface. Professor Young, apparently (though not quite obviously) referring to this point, says, "The meteoric orbit would have to lie nearly in the plane of the sun's equator." But this would not do. Any system of meteors having its perihelion close to the sun's surface, must at the time of the perihelion passage of the flight gather up its members very close together (measuring their distance square to their plane of travel). Some might be a long way ahead of others along the orbit, and the points of nearest approach to the sun might be, some a little nearer, some a little farther from his centre (though even this range of distance must be very small); but the range on one side or the other of the mean plane of the meteoric system, would be exceedingly small at perihelion. For consider, this nearest point is, let us say, half a million miles from the sun's centre, while the farthest point, exactly opposite, would lie, in the case of a system with an eleven-year period, some 900 millions of miles away (or 1,800 times as far); if then two meteors travelling side by side so as to reach perihelion at the same moment, and both at the same distance from the sun, were a thousand miles apart there, they would be 1,800,000 miles apart when simultaneously passing their aphelia. Now, 1,000 miles would be a mere nothing at the sun's distance. His diameter is, roughly, 850,000 miles, and the 850th part of that would be a distance undiscernible by the naked eye, less than the 30th part of the diameter Venus shows when in transit across the sun's disc. Any distance corresponding

'Putting Venus's diameter at 7,500 miles, the circle on the sun's surface hidden by her has a diameter exceeding this as the sun's distance exceeds Venus's, or as twenty-five exceeds seven (roughly): dividing 7,500 thus increased by 850, we get rather more than thirty.

to the actual range of sun-spots in latitude on the sun's surface would correspond to an impossible range of the meteor flight near aphelion: for instance 400,000 miles, corresponding to less than the range from 30° north on the sun to 30° south, would correspond to 720,000,000 miles, in range square to the plane of the orbit, which is, of course, perfectly inconsistent with the idea that the meteors could belong to the same system.

To these difficulties may be added one noted by Professor Young, who points out that it is difficult to make the meteor theory explain the enormous dimensions and persistence of many sun-spot groups, while the irregularity in the epochs of maxima and minima is much greater than would have been expected on this hypothesis.

Yet there are some points in this meteoric and cometic theory of sun-spots which seem so strongly to suggest that we are at least in the track of truth here, that we must not dismiss it hastily because, as presented in one particular way, it seems inconsistent with the observed facts.

And first, be it noticed that the theory of a single meteor system being the cause of the sun-spots was antecedently most unlikely to be true. When we take into account the enormous number of such systems which (as we have shown) must exist in the sun's neighbourhood, the idea that one system of any particular period should have more than others to do with the sun-spot variations, seems altogether incredible at the outset. Even if the eleven-year period were marked with the most perfect regularity this would be so. But as that period is very far indeed from being regularly followed, as the sun-spots wax and wane in number and in size, we might with equal reason reject the single-meteor system theory, from à posteriori considerations. We must admit the probable existence of many sun-disturbing meteor systems, if we adopt the meteoric explanation of the sunspots at all.

Now, it certainly seems a noteworthy circumstance in this connection, that while there are many known comets whose aphelia (or the points of their orbits farthest from the sun) lie near the orbit of Saturn, comets and meteor systems so situated would have about the period which we recognise in the recurrence of solar spots. Putting the greater axis of a comet's elliptic path extending to or a little beyond the orbit of Saturn at ten times the earth's, for convenience of reckoning (it might be somewhat more or less, but ten is a fair estimate enough), we determine the period in which such a comet would circulate around the sun, by Kepler's third law, very readily. Its mean distance is five, and according to that law we have simply

to cube this number (getting 125), and take the square root of the result, getting about eleven and one-fifth, showing that the period of circulation would be eleven and one-fifth years or thereabouts. So that if it so chanced that there were several of these Saturnian meteor systems, whose richest portion reached the sun's neighbourhood at about the same time, there would be something akin to the recurrence of sun-spot maxima; there would also be such variations as are actually observed; and we might readily interpret nearly all the most marked peculiarities of sun-spots so far as the place and time of their appearance on the sun's disc are concerned.

Nay more, it is a noteworthy circumstance that cometic orbits. show a tendency to precisely that degree of inclination to the mean plane of the solar system (in or near which all the planets move), corresponding to the observed position of the sun-spot zones. No comets travel in or near this mean plane, very few travel nearly at right angles to it; the greater number travel on paths inclined between twenty and sixty degrees to that plane. So that their tracks, while near the sun, would be near one or other of the zones where sun-spots chiefly appear.

If, however, we regarded the principal sun-disturbing meteor systems as thus related to the orbit of Saturn, and, adopting a view thrown out long since by myself (without any thought of the theory we are upon), that the Saturnian, Jovian, and going farther from the sun, the Uranian and Neptunian comets and meteor systems were originally expelled from the planet with which they seem thus to be associated, we might find some difficulty in explaining why Saturn rather than Jupiter should seem thus associated with the production of sun-spots. There is no sun-spot period corresponding with the movement of meteor systems to and fro between the sun and Jupiter's orbit, although many comets (and therefore many meteoric systems) exist which have their remotest parts near the orbit of Jupiter. One would be led to expect that as Jupiter is much the larger planet (in fact he surpasses Saturn in mass threefold, and Saturn and all the other planets taken together two and a half times) there would be clear evidence of a sun-spot period of about four and one-third years, the time corresponding to the motion of matter in an orbit having its remotest point near the orbit of Jupiter, and its perihelion close to the sun. But there is no trace of the existence of any such period.

One might perhaps find an explanation of this in the circumstance that Saturn presents all the appearance of being a younger and more active member of the solar system than his brother giant Jupiter.

That ring system of Saturn's, which distinguishes him from all the other planets, is an evidence of extreme planetary youth. He has not yet in fact completed the fashioning of his system. Unlike Jupiter, whose satellite system is complete and symmetrical, Saturn has a system partly incomplete-eight satellites already formed and a system of rings from which other satellites are hereafter to be fashioned. It may well be that while nearly all the comets associated with Jupiter have already done their work, and are now practically eliminated (such of them at least as could effectively disturb the sun), those formed much later by the younger planet still exert a potent influence, and thus still communicate their mean periodic time to the most marked of all the sun-spot periods.

Be this as it may, it certainly is a noteworthy circumstance that the chief sun-spot period should be that which would belong to the Saturnian comets, so to designate those whose orbits have their points of greatest recession from the sun close to the orbit of the distant planet Saturn, the sole member of the solar system which has not yet assumed the form and aspect of an ordinary world, but remains still girt about by a ring system such as science recognises as belonging to the youthful (if one ought not rather to say the embryonic) stages of a planet's existence.

Professor Young, after treating of the meteoric theory of sun-spot formation as one well deserving of careful consideration, says that he shall recur to it; but he does not. Earlier he touches on a meteoric explanation of a certain solar phenomenon well worth considering, though as yet among the mysteries of mysteries which astronomy brings within our ken.

The sun, judged by his photosphere, a visible light-emitting surface, does not rotate all in one, but some parts of him rotate faster than others. His equatorial regions travel fastest, his midlatitudes come next, and the higher latitudes (at least of his cloud envelope) rotate slowest of all. It is easy to take this statement, as presented in the books, without more interest than we should give, perhaps, to the statement that Jupiter rotates in less than ten hours, while the earth takes twenty-four hours to turn once upon his axis. But in reality, this varying rotation rate is a most remarkable phenomenon. Consider what it means. The equatorial parts of the solar surface move round once in about twenty-four days. The surface near the highest solar latitudes where spots have yet been observed, goes round once in about twenty-eight days. Thus, speaking without nice reference to details, the equatorial regions complete seven circuits in 168 days, while in the same time the