Astronomy

Now, if the material of the streamers had been simply a superficial overflow, it should have carried with it into higher latitudes an excess of linear rotational speed, and should hence have pushed its way onwards as it proceeded north and south. But, instead, it fell behind; its velocity was less, not greater than that of the belts with which it eventually became incorporated. What are we to gather from this fact? Evidently that the currents issuing north and south were of eruptive origin. Their motion, in miles per second, was slow, because they belonged to profound strata of the planet’s interior. Their backward drift measured the depth from which they had been flung upward.

The spots, red, white, and black, constantly visible on the Jovian surface, excite the highest curiosity. They are of all kinds and qualities, and their histories and adventures are as diverse as they are in themselves. Some are quite evanescent; others last for years. At times they come in undistinguished crowds, like flocks of sheep, then a solitary spot will acquire notoriety on its own account. White spots appear in both ways; black spots more often in communities; and it is remarkable that the former frequent distinctively, though not exclusively, the southern, the latter the northern hemisphere. Red spots, too, develop pretty freely; but the attention due to them has been mainly absorbed by one striking specimen.

The Great Red Spot has been present with us for at least nineteen years; and it is a moot point whether its beginnings were not watched by Cassini more than two centuries ago. Its modern conspicuousness, however, dates from 1878. Then of a full brick-red hue, and strongly-marked contour, it measured 30,000 by nearly 7,000 miles, and might easily have enclosed three such bodies as the earth. It has since faded several times to the verge of extinction, and partially recovered; but there has never been a time when it ceased to dominate the planet’s surface-configuration. More than once it has been replaced by a bare elliptical outline, as if through an effusion of white matter into a mould previously filled with red matter; and just such a sketch was observed by Gledhill in 1870. The red spot is attached, on the polar side, to the southern equatorial belt. It might almost be described as jammed down upon it; for a huge gulf, bounded at one end by a jutting promontory, appears as if scooped out of the chocolate-coloured material of the belt to make room for it. Absolute contact, nevertheless, seems impossible. The spot is surrounded by a shining aureola, which seemingly defends it against encroachments, and acts as a chevaux de frise to preserve its integrity. The formation thus constituted behaves like an irremovable obstacle in a strong current. The belt-stuff encounters its resistance, and rears itself up into a promontory or “shoulder,” testifying to the solid presence of the spot, even though it be temporarily submerged. The great red spot, the white aureola, and the brownish shoulder are indissolubly connected.

The spot is then no mere cloudy condensation. Yet it has no real fixity. Its period of rotation is inconstant. In 1879–80, it was of 9 hours, 55 minutes, 34 seconds; in 1885–86, it was longer by 7 seconds. The object had retrograded at a rate corresponding to one complete circuit of Jupiter in six years, or of the earth in seven months.^[58] It is not then fast moored, but floats at the mercy of the currents and breezes predominant in the strange region it navigates. A quiescent condition is implied by the approximate constancy of its rotation-period during the last ten years. With the paling of its colour, its “proper motion” slackens or ceases. This must mean that, at its maxima of agitation, it is the scene of uprushes from great depths, which, bringing with them a slower linear velocity, occasion the observed laggings. It is not self-luminous, and shows no symptom of being depressed below the general level of the Jovian surface. A promising opportunity was offered in 1891 of determining its altitude relative to a small dark spot on the same parallel, by which, after months of pursuit, it was finally overtaken. An occultation appeared to be the only alternative from a transit; yet neither occurred. The dark spot chose a third. It coasted round the obstacle in its way, and got damaged beyond recognition in the process. Its material, as Mr. Stanley Williams observed, “was diverted and forced bodily southwards, and obliged to pass round the southern side of the red spot as if it were an island projecting above a stream.”

Jupiter has no certain and single period of rotation. Nearly all the spots that from time to time come into view on its disc are in relative motion, and thus give only individual results. The great red spot has the slowest drift of all (with the rarest exceptions), while the black cohorts of the northern hemisphere outmarch all competitors. Mr. Stanley Williams,^[59] as the upshot of long study, has delimitated nine atmospheric surfaces with definite periods. They are well marked, and evidently have some degree of permanence, yet the velocities severally belonging to them are distributed with extreme irregularity. Thus, two narrow, adjacent zones differ in movement by 400 miles an hour. This state of things must obviously be maintained by some constantly acting force, since friction, if unchecked, would very quickly abolish such enormous discrepancies. The rotational zones are unsymmetrically placed; there is no correspondence between those north and south of the Jovian equator; and, although the equatorial drift is quicker than that of either tropic, it is outdone in 20° to 24° north latitude. The stability of this anomalous mode of rotation was remarkably illustrated by Dr. Rambaud’s measurements of the “Garnet Spot” of October, 1895. Its movement proved to be strictly conformable to that of the zone in which it was situated (10° to 20° north latitude), and to agree, moreover, within a fifth of a second with the value deduced by Schröter in 1787 for that of a spot in the same “zenographical” district.^[60]

Jupiter’s equatorial rotation, as indicated by observations of spots, is accomplished in 9 hours 50 minutes; but Bélopolsky’s and Deslandres’ spectrographic determinations gave rates of approach and recession falling somewhat short of the corresponding velocity.^[61] Possibly the spots forge ahead in the medium that sustains them; or it may be, as M. Bélopolsky suggests, that the planetary sphere itself has been measured too large, owing to refraction in its atmosphere.

However this be, the rotation of the great planet, albeit ill-regulated (if the expression be permissible), is distinctly of the solar type. It is itself a “semi-sun,” showing no trace of a solid surface, but a continual succession of cloud-like masses belched forth from within. Each series, in fact, of certain classes of markings, such as the equatorial “port-holes,” plainly owes its origin to the rhythmical activity of a solitary, deep-buried focus.^[62] Jupiter’s low mean density, considered apart from every other circumstance, suffices to demonstrate the primitive nature of his state. Under the enormous pressure reigning in his interior, the same materials should be vastly more massive, specifically, than within our own small globe; their fourfold expansion gives us to understand the intensity of that heat by which pressure has been so much more than neutralised. Moreover, the agitations due to the cooling of a fluid globe make their mark on its turbulent surface. On a solidified body like the earth, circulation is kept up by heat received from without, and is purely atmospheric, and essentially horizontal. In a sun-like body, the circulation is bodily and vertical. That the processes going on in Jupiter are of this kind is beyond question. Exchanges of hot and colder substances are effected, not by surface-flows, but by up and down rushes. The parallelism of his belts to his equator makes this visible to the eye. An occasional oblique streak^[63] betokens a current in latitude, but it is exceptional, and might be called out of character.

Jupiter’s true atmosphere encompasses the disturbed shell of vapours observed telescopically. Its general absorptive action upon light is betrayed by the darkening of the planet’s limb—another point of resemblance to the sun; while its special, or selective, absorption can only be detected with the spectroscope. The arresting effect of water-vapour was early noticed by Huggins and Vogel, and they measured a strong line in the red of unknown origin, but contained in banded star spectra. Atmospheric absorption is strongest above the ruddy equatorial belts, which are hence concluded to be placed at a lower level than the white surface.

Planetary photography was set on foot by Dr. Gould of Boston, in 1879, when he obtained some promise of success with Mars, Jupiter, and Saturn; and Dr. Lohse prosecuted the subject in 1883. The actinic power of Jupiter’s light is very remarkable. It surpasses that of moonlight nine times, and that of Mars twenty-four times. Dr. Lohse further ascertained that the southern hemisphere is twice as chemically effective as the northern.^[64] This superiority is doubtless connected with the greater physical agitation of the same region. A series of photographs of Jupiter, taken in 1891 with the great Lick refractor, were the first of any value for purposes of investigation. Each is one inch in diameter; the image of the planet having been enlarged eight times before being received upon the plate. Mr. Stanley Williams found them full of interesting detail. Figure 16 shows an enlargement of a striking photograph taken by Professor E. C. Pickering.

Jupiter’s satellites were the first trophies of telescopic observation. They are, indeed, bright enough for naked eye perception, could they be removed from the disc which obscures them with its excessive splendour; and the first and third have actually been seen, in despite of the glare, by a few persons with phenomenally good eyesight. The mythological titles of the Galilean group—Io, Europa, Ganymede, and Calypso (proceeding from within outward) have been superseded by prosaic numbers. The change was unlucky, but is now probably irremediable.

The Jovian family presents an animated and attractive spectacle. The smallest of its original members (No. II.) is almost exactly the size of our moon; the largest (No. III.), with its diameter of 3,550 miles, considerably exceeds the modest proportions of Mercury. Satellite I. revolves in 42½ hours at the same average distance from Jupiter’s surface that our moon does from that of the earth. No. II. has a period of 3 days 13 hours, and its distance from Jupiter’s centre is 415,000 miles. Both these orbits are sensibly circular; and Nos. III. and IV. travel in ellipses of very small eccentricity, the one at a mean distance of 664,000, the other at 1,167,000 miles, in periods respectively of 7 days 4 hours, and 16 days 16½ hours. All four revolve strictly in the plane of Jupiter’s equator.

They constitute a system bound together by peculiar dynamical relations, in consequence of which they can never be all either eclipsed, or seen aligned at one side of their primary, at the same time. They can all, however, be simultaneously hidden behind it, or in its shadow; although this moonless condition is looked out for as a telescopic rarity.

The varied phenomena of eclipses, occultations, and transits, offer the interest, not only of predictions fulfilled, but sometimes of discrepancies detected. The three inner satellites plunge through the huge neighbouring shadow-cone at every revolution; the fourth, owing to its greater distance, escapes eclipse when the shadow makes an appreciable angle with the plane of its orbit. When Jupiter is in opposition or conjunction, occultations, but no eclipses, of his moons take place; at other periods, the two kinds of obscuration merge into, or succeed each other. “Time cannot stale their infinite variety.”

From observations of the eclipses of Jupiter’s satellites, Olaus Römer gathered, in 1675, the first intimations of the finite velocity of light. He noticed that their visibility was alternately retarded and accelerated as the earth withdrew from, and approached the scene of their occurrence; and he designated half the extreme difference, or the time occupied by light in travelling from the earth to the sun, the “equation of light.” Its value is 500 seconds; and until recently, no other measure was available of that fundamental constant of nature—the rate of luminous transmission.

The transits of the satellites across the Jovian disc present many curious appearances, due to complicated and changeable effects of light and shade both upon the planetary background, and upon the little circular objects self-compared with it. These, in the ordinary course, show bright while near the dusky limb, then vanish during the central passage, and re-emerge again bright at the opposite side. But, instead of duly vanishing, they now and then darken even to the point of becoming indistinguishable from their own shadows, by which they are preceded or followed. This difference of behaviour cannot be attributed wholly to varieties of lustre in the sections of the disc transited; otherwise, it could be predicted. But this has never been attempted; “black transits” come when least expected. The third and fourth satellites are those chiefly subject to these phases; the second has never been known to exhibit them; and they but slightly affect the first. A drawing by Professor Barnard of one of its bright transits with an attendant shadow that Peter Schlemyl might have envied, is reproduced in Figure 17. Its belted appearance, detected by that eminent observer, will be noted. Indeed, all the satellites, except perhaps No. II. are striped or spotted; and this leads to seeming deformations in their shape, as well as fluctuations in their brightness, the markings being evidently of atmospheric origin, and hence changeable. Their distinct and accurate perception has been made possible by the excellence of the Lick thirty-six inch refractor.

Jupiter’s moons seem to resemble him in constitution. The three first possess the same high reflective power. No. II. is as bright as the planet’s brightest parts, so that its albedo cannot fall short of 0·70. And even No. IV. (formerly designated “Calypso” in reference to its frequent obscurations) exactly matches, during its darkest phases, the blue-grey polar hoods of its primary. On an average, too, the satellites seem to be of about the same mean density as Jupiter, No. I. being considerably the lightest for its bulk; and their spectra, according to Vogel’s observations in 1873, are composed of solar rays modified in precisely the same way as those reflected by the planet. Nothing is known quite certainly about their rotation-periods. Sir William Herschel concluded them to be of the same length with their periods of revolution; but recent work throws some doubt upon the reality of this agreement.

The discovery, September 9, 1892, of Jupiter’s “fifth satellite” was one of the keenest astronomical surprises on record. An accession to a system so symmetrically arranged, so complete, to our judgment, as it stood, appeared superfluous, and, considering the eager scrutiny devoted to it during 282 years, well-nigh incredible. But the extra member was in truth out of reach until it was found; original discovery being, as every one knows, a greatly more arduous feat than subsequent verification. Nor could it have been casually detected. Professor Barnard seized the opportunity, lent by the specially favourable opposition of 1892, to rummage the system for novelties. Keeping the telescopic field dark by means of a metallic bar placed so as to occult the gorgeous planetary round, he sought, night after night, for what might appear. At length, on September 9, he caught the glimmer he wanted, and made sure, September 10, that it truly intimated the presence of a new satellite.

This small body revolves in a period of 11 hours, 57 minutes, 23 seconds, at a mean distance of 112,160 miles from Jupiter’s centre, or 67,000 from his bulged equatorial surface. Hence, it should by right be called “No. I.” instead of “No. V.” The major axis of the ellipse in which it circulates advances so rapidly, owing to the disturbance caused by Jupiter’s spheroidal figure, as to complete a revolution in five months. The implied eccentricity of its orbit, as M. Tisserand has shown,^[65] very slightly exceeds that of the orbit of Venus, yet it has been made obvious by Barnard’s observations of the differences between its east and west elongations. Its orbital velocity of 16½ miles a second far surpasses that of any other satellite in the solar system. Close vicinity to a mass so vast as Jupiter’s demands counter-balancing swiftness. Its period of revolution being, however, longer by one hour than Jupiter’s period of rotation, it so far conducts itself normally as to rise in the east and set in the west. On the other hand, since its progress over the sphere is measured by the difference between the two periods, it spends five Jovian days in journeying from one horizon to the other, running, in the meantime, four times through all its phases. Yet it never appears full. Jupiter’s voluminous shadow cuts off sunlight from it during nearly one-fifth of each circuit.

It is an exceedingly elusive telescopic object. There is no chance of catching a glimpse of it except with a powerful and perfect telescope at its “elongations,” or furthest excursions of about eight seconds of arc on either side of the planet For the most part, it lurks within the blaze as closely as Teucer behind the shield of Ajax. It is far too small to be discerned in projection upon the disc, which, viewed from it in mid-transit, is full with a diameter of 42° 2′, and an area 6,440 times that of our moon. Yet, since its intrinsic lustre is less in the proportion of 2 to 15, the light shed by Jupiter upon the “fifth satellite” equals the joint radiance of no more than 860 full moons.

The new satellite is indistinguishable in aspect from a star of the thirteenth magnitude. And its neighbour No. I. being of 5·6 magnitude, we receive from it 910 times more light than from the stranger. If both be equally reflective, the diameter of the latter is ¹⁄₃₀th the diameter of the former, or, approximately, 80 miles. But its albedo is unlikely to exceed that of Mars. By a rough estimate, therefore, this interesting object measures 120 miles across, and 9000 such miniature globes would go to the making of one full-sized Jovian attendant. Instead of being a late addition to the system, or, so to speak, an afterthought, it may be presumed, from the perceptible eccentricity of its path, to be the senior member of the family. But the subject of its origin is not yet ripe for discussion.

Nearly twice as far from the sun as Jupiter revolves a planet, the spacious orbit of which was, until 1781, supposed to mark the uttermost boundary of the solar system. The mean radius of that orbit is 886 millions of miles; but in consequence of its eccentricity, the sun is displaced from its middle point to the extent of 50 million miles, and Saturn is accordingly 100 million miles nearer to him at perihelion than at aphelion. The immense round assigned to the “saturnine” planet is traversed in 29½ years, at the tardy pace of six miles a second. His seasons are thus twenty-nine times more protracted than ours, and are nominally more accentuated, since his axis of rotation deviates from the vertical by 27°. But solar heat, however distributed, plays an insignificant part in his internal economy. In the first place, its amount is only ¹⁄₉₁th its amount on the earth; in the second, Saturn, like Jupiter—even more than Jupiter—is thermally self-supporting. The bulk of his globe comparatively to its mass suffices in itself to make this certain. The mean diameter of Saturn is 71,000 miles, or nine times (very nearly) that of the earth; if of equal density, its mass should then be nine cubed, or 729 times the same unit The actual proportion, however, is 95; hence the planet has a mean density of only ⁹⁵⁄₇₂₉, or between ⅐th and ⅛th the terrestrial, and being thus composed of matter as light as cork, would float in water. Professor G. H. Darwin has moreover demonstrated, from the movements of its largest satellite, that its density gains markedly with descent into the interior, so that its surface-materials must be lighter than any known solid or liquid.

When at its nearest to the earth, Saturn is as large as a sixpence held up at a distance of 210 yards.^[66] But instead of being round like a sixpence, it is strongly compressed—more compressed even than Jupiter. The spectra of the two planets are almost identical. Both are impressed with traces of aqueous absorption, and include the “red star line.” About the albedo of Saturn there is some uncertainty. Zöllner made it 0·50, a very probable value; Müller of Potsdam determined it at 3·3 times that of Mars, the unit of his scale. For the value of the unit, the only authority is Zöllner, who found Mars to give back 0·26 of the light dispensed to him. Multiplying then 0·26 by 3·3 we get for the albedo of Saturn 0·86, an impossible number for a non-luminous body, the albedo of “untrodden snow” being, as already stated, 0·78.

Saturn resembles to the eye a large, dull star; its rays are entirely devoid of the sparkling quality which distinguishes those of Jupiter. But it shows telescopically an analogous surface-structure. Its most conspicuous markings are tropical dark belts of a greyish or greenish hue; the equatorial region is light yellow, diversified by vague white spots; while the poles carry extensive pale blue canopies. The apparent tranquillity of the disc may be attributed in part to the vast distance from which it is viewed; yet not wholly. For lack of fiducial points, no attempt was made to determine the planet’s rotation until 1794, when the elder Herschel, by following an identified irregularity in a complex banded formation, arrived at a period of 10 hours 16 minutes. The first possibility of checking this result offered itself to Professor Hall of Washington, after fourteen years of vain expectation, in the emergence of a white spot just north of the equator, the movement of which gave for the length of the Saturnian day, 10 hours, 14 minutes, 24 seconds. In 1891–2, Mr. Stanley Williams made observations upon a good many such objects; and their discussion by Mr. Denning afforded a mean period two seconds longer than Hall’s. Individual variations, however, to the extent of 14 seconds were brought out by it, proving that Saturnian, like Jovian, spots have “proper motions,” and cannot be depended upon to give the true rotation of the planet. Its compound nature may be suspected, but has not yet been proved.

From measures executed by Barnard in 1895, it appears that the equatorial diameter of Saturn is 76,470, its polar diameter 69,770 miles, giving a mean diameter of 74,240, and a compression of about ¹⁄₁₂. Gravity, at its surface, is only one-fifth more powerful than on the earth.

Thus, Saturn not only belongs to the same celestial species as Jupiter, but is a closely-related individual of that species. There is no probability that either is to any extent solid. Both exhibit the same type of markings; both betray internal tumults by eruptions of spots which, by their varying movements, supply a measure for the profundity of their origin; both possess identically constituted atmospheres, and are darkened marginally by atmospheric absorption.

Saturn is, however, distinguished by the possession of an unique set of appendages. Nothing like them is to be seen elsewhere in the heavens; and when well opened (as in Fig. 18) they form, with the globe they enclose, and the retinue of satellites in waiting outside, a strange and wonderful telescopic object. The rings, since they lie in the plane of Saturn’s equator, are inclined 27° to the Saturnian orbit, and 28° to the ecliptic. The earth is, however, comparatively to Saturn, so near the sun, that their variations in aspect, as viewed from it, may in a rough way be considered the same as if seen from the sun. They correspond exactly with the Saturnian seasons. At the Saturnian equinoxes, the rings are illuminated edgewise, and disappear, totally or approximately; at the Saturnian solstices, sunlight strikes them nearly at the full angle of 27°, first from below, then from above. At these epochs, we perceive the appendage expanded into an ellipse about half as wide as it is long. Two concentric rings (generally called A and B) are then very plainly distinguishable, the inner being the brighter. The black fissure which separates them is called “Cassini’s division,” because that eminent observer was, in 1675, the first to perceive it. A chasm known as “Encke’s division,” in the outer ring (A), is a thinning out rather than an empty space; and temporary gaps frequently appear in A, while B is entirely exempt from them. There are then two definite and permanent bright rings, and no more; but with them is associated the dusky formation discovered by W. C. Bond, November 15, 1850, and described by Lassell as “something like a crape veil covering a part of the sky within the inner ring.” It is semi-transparent the limb of Saturn showing distinctly through it.

The exterior diameter of the ring-system is 172,800, while its breadth is 42,300 miles.^[67] The rings A and C are each 11,000 miles wide; while B measures 18,000, Cassini’s division 2,270, and the clear interval between C and the planetary surface somewhat less than 6,000 miles. Each ring, C included, is brightest at its outer edge; but there is no gap between the shining and the dusky structures, B shading by insensible gradations up to C, yet maintaining distinctness from it. The earliest exact determinations of the former were made by Bradley in 1719, since when they have been affected by no appreciable change.^[68] The theoretically inevitable subversion of the system is progressing with extreme slowness.

The thickness of the rings is quite inconsiderable. They are flat sheets, without (so to speak) a third dimension. For this reason, they disappear utterly in most telescopes, when their plane passes through the earth, as it does twice in each Saturnian year. Only under exceptional conditions, a narrow, knotted, often nebulous, streak survives as an index to their whereabouts. On October 26, 1891, Professor Barnard,^[69] armed with the Lick refractor, found it impossible to see them projected upon the sky, notwithstanding that their shadow lay heavily on the planet It was not until three days later, that “slender threads of light” came into view. The corresponding thickness of the formation was estimated at less than fifty miles. The phenomenon of the disappearance of the rings will not recur until July 29, 1907.

The constitution of this marvellous structure is no longer doubtful. It represents what might be called the fixed form of a revolving multitude of diminutive bodies. This was demonstrated by Clerk Maxwell in the Adams Prize Essay of 1857. His conclusion proved irreversible. The pulverulent composition of Saturn’s rings is one of the acquired truths of science. An incalculable number of tiny satellites, revolving independently in distinct orbits, in the precise periods prescribed by their several distances from the planet, are aggregated into the unmatched appendages of Galileo’s tergeminus planeta. The local differences in their brightness depend upon the distribution of the component satelloids. Where they are closely packed, as in the outer margins of rings A and B, sunlight is copiously reflected; where the interspaces are wide, the blackness of the sky is barely veiled by the scanty rays thrown back from the thinly scattered cosmic dust. The appearance of the crape ring as a dark stripe on the planet results—as M. Seeliger has pointed out—not from the transits of the objects themselves, but from the flitting of their shadows in continual procession across the disc.

The albedo of these particles is so high as to render it improbable that they are of an earthy or rocky nature, such as the meteorites which penetrate our atmosphere. The rings they form are, on the whole, more lustrous than Saturn’s globe; but this superiority is held to be due to the absence of atmospheric absorption. Their spectrum is that of unmodified sunlight.

An eclipse of Japetus, the eighth Saturnian moon, by the globe and rings, November 1, 1889, was highly instructive as to the nature of the dusky appendage. The satellite was never lost sight of during its passage behind it; but became more and more deeply obscured as it travelled outward; then, at the moment of ingress into the shadow of ring B, suddenly disappeared. Certainty was thus acquired that the particles forming the crape ring are most sparsely strewn at its inner edge—which is, nevertheless, perfectly definite—and gradually reach a maximum of density at its outer edge. Yet, while there is not the smallest clear interval, a sharp line of demarcation separates it from the contiguous bright ring. Professor Barnard was the only observer of these curious appearances. The distribution of the ring-constituents, like that of the asteroids, was governed by the law of commensurable periods, Saturn’s moons replacing Jupiter as the perturbing and regulating power. Kirkwood showed in 1867, that Cassini’s division represents a region of peculiarly strong disturbance; since a body revolving there would have a period connected by a simple relation with the periods of no less than four satellites. Encke’s division, too, as Dr. Meyer has indicated, and other lines of scanty occupation and occasional vacancy, coincide with districts of space where similar combinations occur.

The “satellite-theory” of Saturn’s rings has received confirmation from apparently the least promising quarters. Professor Seeliger of Munich showed, from photometric experiments in 1888, that their constant lustre under angles of illumination ranging from 0° to 30° was proof positive of their composition out of discrete small bodies.^[70] And Professor Keeler of Alleghany, by a beautiful and refined application of the spectroscopic method, arrived at the same result in April, 1895.^[71] “Under the two different hypotheses,” he remarked, “that the ring is a rigid body, and that it is a swarm of satellites, the relative motion of its parts would be essentially different.” The former would necessarily involve increasing velocity outward, the latter, increase of velocity inward, just for the same reason that Mercury moves more swiftly than the earth, and the earth than Saturn; while the sections of a solid body, which could have but one period of rotation, should move faster, in miles per second, the farther they were from the centre of attraction. The line of sight test is then theoretically available; but it was an arduous task to render it practically so. The difficulties were, however, one by one overcome; and a successful photograph of the spectra of Saturn and its rings gave the required information in unmistakable shape. From measurements of the inclinations of five dusky rays contained in it with reference to a standard horizontal line, rates of movement were derived of 12½ miles per second for the inner edge of ring B, and of 10 miles for the outer edge of ring A. The agreement with theory was, as nearly as possible, exact; the components of the rings were experimentally demonstrated to be moving, each independently of every other, under the dominion of Kepler’s laws.

For the globe of Saturn, Professor Keeler obtained, by the same exquisite method, a rotational period of 10 hours, 14 minutes, 24 seconds, in precise accordance with that indicated by the white spot of 1876, which thus seems to have had no proper motion, but to have floated on the ochreous equatorial surface as tranquilly as a water-lily upon a stagnant pool. The result, so far as it goes, hints that Saturn may be really, as well as apparently, less ebullient than Jupiter.

Seers into the future of the heavenly bodies consider that the rings of Saturn, like the gills of a tadpole, are symptomatic of an early stage of development; and will be disposed of before he arrives at maturity. They cannot be regarded otherwise than as abnormal excrescences. No other planet retains matter circulating round it in such close relative vicinity. It was proved by Roche of Montpellier that no secondary body of importance can exist within less than 2·44 mean radii of its primary; inside of that limit, it would be rent asunder by tidal strain. But the entire ring-system lies within the assigned boundary; hence, being where it is, it can only exist as it is—in flights of discrete particles. Will it, however, always remain where it is?

“Clerk Maxwell,” wrote Mr. Cowper Ranyard,^[72] “used to describe the matter of the rings as a shower of brickbats, amongst which there would inevitably be continual collisions. The theoretical results of such impacts would be a spreading of the ring both inwards and outwards. The outward spreading will in time carry the meteorites beyond Roche’s limit, where, in all probability, they will, as Professor Darwin suggests, slowly aggregate, and a minute satellite will be formed. The inward spreading will in time carry the meteorites at the inner edge of the ring into the atmosphere of the planet, where they will become incandescent, and disappear as meteorites do in our atmosphere.”

Yet it may be that collisions are infrequent in this conglomeration of “brickbats.” There is the strongest presumption that they all circulate in the same direction, in orbits nearly circular, and scarcely deviating from the plane of the Saturnian equator. Those pursuing markedly eccentric tracks must long ago have been eliminated. Thus, encounters can only occur through gravitational disturbances by Saturn’s moons, and they must be of a mild character, depending upon very small differences of velocity. The first sign of a “spreading outwards” should be the formation of an exterior “crape ring,” of which no faintest trace has yet been perceived.

Saturn’s rings are entirely invisible from its polar regions, but occasion prolonged and complex eclipse-effects in its temperate and equatorial zones. They have been fully treated of from the geometrical point of view by Mr. Proctor in “Saturn and its System.”

Of this planet’s eight satellites, the largest, Titan (No. VI.), was discovered first (by Huygens in 1655), and the smallest, Hyperion (No. VII.), last (by Lassell and Bond in 1848). The five others were detected by J. D. Cassini and William Herschel. Titan, alone of the entire group, equals our moon in size. It measures, according to Professor Barnard, 2,720 miles across. Its period of revolution is nearly sixteen days, its distance from Saturn’s centre, 771,000 miles. The orbit of Japetus (No. VIII.) is the largest, and its period the longest of any secondary body in the solar system. It circulates in 79⅓ days at a distance of 2,225,000 miles, equal to 59½ of Saturn’s equatorial radii. Hence its path is of about the same proportional dimensions as that of our moon. Japetus is remarkable for its variability in light. It is capable of tripling or quadrupling its minimum lustre. Sir William Herschel noticed that these maxima coincided with a position on the western side of the planet, and inferred rotation of the lunar kind. “From the changes in this body,” he argued in 1792,^[73] “we may conclude that some part of its surface, and this by far the largest, reflects much less light than the rest; and that neither the darkest nor the brightest side is turned towards the planet, but partly one and partly the other, though probably less of the bright side.”

This explanation, however, he admitted to be incomplete. There was, and is, outstanding variability, which seems to intimate the presence of an atmosphere and the formation of clouds. But no positive knowledge has yet been gained regarding the physical state of Saturn’s moons. We may nevertheless conjecture that, since tidal friction has destroyed the rotation (as regards Saturn) of the remotest member of the family, it has not spared those more exposed to its grinding-down action. All presumably rotate in the same time that they revolve.

The five inner satellites move in approximately circular orbits; the three outer in ellipses about twice as eccentric as the terrestrial path. All, Japetus only excepted, keep strictly to the plane of the rings. And since this makes an angle of 270 with the planet’s orbit, eclipses are much less frequent here than in the Jovian system. They can only occur when Saturn is within a certain distance (different for each) from the node of the satellite-orbit. Even Mimas (No. I.), although it wheels round the ring at an interval of only 34,000 miles, often slips outside the obliquely-projected shadow-cone. Its distance from Saturn’s centre is 118,000 miles, and it completes a circuit in 22½ hours. Perpetually wrapped in the glare of its magnificent primary, it is a very shy object, only to be caught sight of in its timid excursions by the very finest telescopes. Like all the Saturnian moons, except Titan, and, by a rare conjuncture, Japetus, it is far too much contracted to be visible in transit across the disc.

The movements of these bodies have been carefully studied, and their mutual perturbations to some extent unravelled. They have proved exceedingly interesting to students of celestial mechanics. Titan has, in this department, chiefly to be reckoned with. He exercises in the Saturnian system a similar overpowering influence to that wielded by Jupiter in the solar system. Mr. Stone finds its mass to be ¹⁄₇₆₀₀th that of Saturn, showing that its density is nearly equal to that of our moon. This seems to indicate an advanced stage of cooling. On the other hand, its albedo is evidently very high. The other satellites appear in the largest telescopes as mere stellar points.

The four giant planets, closely allied as they are, and strongly distinguished in physical constitution from the terrestrial planets, divide again of themselves into two sub-groups. Jupiter and Saturn have much more in common than either has with Uranus or Neptune; while Uranus and Neptune present peculiar analogies. Conclusions concerning one may almost be said to apply to the other. Their enormous distance, it is true, tends to efface minor differences; yet it is insufficient to obliterate similarities of a peculiar kind.

Uranus is a globe 32,000 miles in mean diameter, and decidedly elliptical in shape. Mädler and Schiaparelli agreed in assigning to it a compression of ¹⁄₁₁; Barnard, in 1894, uninformed of their results, noticed the disc to be more oval than Saturn’s. The indicated rotational movement must be very swift; and a lucid spot watched by MM. Perrotin and Thollon at Nice in 1884, seemed to fix it at about ten hours. This was, however, only a vague estimate. Faint equatorial belts, too, have with difficulty been seen. Remembering, indeed, that the object they diversify is just large enough to be annularly eclipsed by a cricket ball two miles off, there is little cause for surprise at the indistinctness of its surface-markings. They probably consist, like those of Jupiter and Saturn, in dusky polar hoods, a brilliant equatorial zone, and obscure intermediate bands. The last were seen as “the merest shades on the planet’s surface,” and under a somewhat deformed aspect, by the Lick observers in 1890 and 1891.^[74] By Professor Young in 1883, on the other hand, and by the MM. Henry at Paris in 1884, they were observed to be symmetrically placed, parallel one to the other, and of what might be called the normal type for great planets. That they constitute, with the bright space they enclose, an equatorial scheme of marking, was proved by Barnard’s comparison of the trend (or position angle), determined for them by Young, with the direction of the shortest axis of the little disc they traverse.^[75] Their considerable foreshortening in 1894 was, doubtless, the reason why Barnard, with his acute vision, was compelled to rely upon earlier observations, brought up to date by computation. Unless, indeed, the markings are intrinsically variable.

This was suspected at Nice in 1889, when a thirty-inch refractor was available for their scrutiny.^[76] Dusky rulings were obvious on a strongly compressed spheroid; and they ran parallel to the major axis of the spheroid—that is, to the planet’s equator. But their appearance varied, and their width seemed irregular. At the same establishment, but with a fourteen-inch telescope, Uranus was observed, under particularly favourable circumstances, March 18, 1884.^[77] An unexpected resemblance to Mars was apparent. The ordinarily sea-green disc was divided into a sombre north-western and a bluish-white south-eastern hemisphere. Dark spots were visible, and a conspicuous white one at the limb simulated a snow-cap. But ulterior observations resolved the spots into belts, and showed the shining patch to be, not polar, but equatorial. It was presumably of an eruptive nature.

The axis upon which Uranus rotates is very much bowed towards the plane of its orbit. Its seasons are hence abnormal; but their vicissitudes can scarcely be sensible at a distance from the sun more than twice that of Saturn. This, as Mr. Proctor noticed, is the only case in which the ratio of one to two is exceeded in the radii of two adjacent planetary orbits. The radius of the Uranian track, pursued at the leisurely pace of 4⅕ miles a second, is 1,782 millions of miles, or more than 19 astronomical units. It consequently receives from the sun 370 times less warmth and light than the earth does. Area for area, it is true, the sun shines with the same intensity there as here; the difference lies in its apparent size. Instead of the broad eye of day to which we are accustomed, the luminary of Uranus presents a surface only 2¼ times that of Jupiter, as seen from the earth at an unfavourable opposition; and although Uranus is 166 millions of miles nearer to the sun at perihelion than at aphelion, no conspicuous difference would mark the passage from one to the opposite point. This is accomplished in 42, the entire round in 84 years.

In point of size, as Professor Young remarks, Uranus compares with the earth very much as the earth compares with the moon. For its surface exceeds the terrestrial surface about sixteen times, and its volume amounts to sixty-six times the terrestrial volume. Its mass, however, is less than fifteen times that of the earth, whence its density is represented (in round numbers) by the fraction ¹⁵⁄₆₆. The large globe is then nearly five times less dense than the small one, its materials exceeding the weight of an equal bulk of water by only one-fifth. Gravity is actually less at its surface than at the sea-level on the earth. Every ton of coal, for instance, delivered in that remote globe would fall short by two hundred pounds. The albedo of Uranus differs little from that of Jupiter; if anything, it is somewhat higher, and is nearly represented by the brilliancy of white paper.

The spectrum of Uranus indicates an emphatic departure from the planetary conditions so far met with. This body is obviously surrounded by a powerfully absorptive atmosphere, of a constitution foreign to our experience. The greenish hue of the light which has traversed some of its strata gives a preliminary indication of the manner in which it has been affected. This its spectrum, first inspected by Secchi in 1869, expounds in detail. He noticed a number of heavy dark bands in the red, while the green and blue sections remaining open gave to the planet its characteristic colour. A couple of years later, Huggins and Vogel executed concordant measurements of six pronounced bands, besides some faint streaks; and on June 3, 1889, the former obtained, with two hours’ exposure, a beautiful spectrographic impression extending far up into the ultra-violet. A corroborative, though less comprehensive, photograph was taken by Mr. Frost at Potsdam, April 23, 1892. Both included many Fraunhofer lines, the presence of which demonstrates that the light of Uranus, although more powerfully stamped with original absorption than that of the rest of the planets, consists essentially of reflected solar rays. Professor Keeler’s admirable series of visual observations with the Lick refractor were undertaken in 1889 to test the truth of a suggestion that this peculiar spectrum consisted of bright bands upon a dark ground, and not of dark bands upon a bright ground. His decision in favour of the latter alternative was without appeal.

Of the six principal dark bands representing the arresting action upon light of the planetary atmosphere, four are quite distinctive; the fifth is the “red star line” common to the spectra of Jupiter and Saturn; the sixth is the hydrogen “F” (Hβ)—not definite and narrow as it is seen in the solar spectrum, but hazy, and graduating in darkness towards the middle, an undoubted outcome of native absorption.^[78] Now, this is a fact that implies a great deal. It gives direct evidence of a very high temperature. Free hydrogen ceases to be present in a body upon which water can form—given, of course, the presence of oxygen, which it would be in the highest degree arbitrary to exclude. At one epoch of its development, the earth must have been surrounded by immense volumes of hydrogen. But with the diminution of heat, union with oxygen became possible, and the gas vanished to reappear in the form of liquid oceans, with their related hydrographic and cloud-systems. Uranus is presumably—almost certainly—still too hot to permit the combination of hydrogen and oxygen; and the absence from its spectrum of the slightest trace of aqueous absorption strengthens this inference. Doubtless, the time will come when the two elements will no longer be held at arms’ length; their affinities will come into play; the familiar, all-important terrestrial liquid will be formed, and the geological history of Uranus will begin.

Uranus is attended by four moons. They are named Ariel, Umbriel, Titania and Oberon. Titania—the third in order of distance from the primary—is the brightest of the group, and has a diameter of possibly one thousand miles. Oberon is slightly inferior. Both were detected by Herschel in 1787. Ariel and Umbriel, captured by Lassell at Malta in 1851, are insignificant bodies in themselves—their dimensions probably differing but slightly from those of Hyperion, the seventh and least Saturnian moon, estimated to measure five hundred miles across. They are among the most difficult of telescopic objects, since they circulate about as close to Uranus as Mimas and Enceladus do to Saturn, are physically smaller, and more than twice as remote from the earth. Both were believed variable by Lassell, and Newcomb obtained in 1875 plausible, though not convincing, evidence that Ariel, at any rate, is subject to light changes in the period of its orbital circulation, showing that, here again, tidal friction has done its work of synchronising rotation and revolution.^[79] None of the four orbits are appreciably eccentric; they all lie in the same plane, and are described in periods ranging from 2½ to 13½ days.

The position of that plane is, however, exceedingly remarkable. It is tilted at an angle of 98° to the ecliptic. This means that the satellites move backward, against the succession of the zodiacal signs. For direct becomes retrograde motion automatically, so to speak, by turning the plane in which it is performed beyond the limit of the vertical. The same fact is merely expressed in two different ways by saying that the bodies in question travel from west to east at an angle of 98°, or from east to west at an angle of 82° to the ecliptic. The planes of the ecliptic and of the Uranian orbit deviate, it should be mentioned, by only two-thirds of a degree. The disturbance by which the Uranian system was set topsy-turvy did not in the least affect the motion of Uranus itself.

Another unusual circumstance about that system is that the satellite-plane departs widely from the equatorial plane. Our own moon, it is true, is similarly circumstanced; but, on the Uranian scale, it is nearly eight times farther from its primary than Ariel, and 2·6 times farther than Oberon; while the enormous equatorial protuberance of Uranus almost seems to impose conformity upon bodies revolving so close to it. Conformity, none the less, is absent. The direction taken by the equator of Uranus, as we have seen, is indicated in a two-fold manner: first, by the trend of the belts; secondly, by the lie of the major axis. And these indications agree. Supposed discrepancies between them have been reconciled by improvements in the conditions of observation. But with the equatorial line the plane of satellite-revolution cannot be brought to coincide. The angle of divergence is uncertain, but may be put roughly at 20°. This would give 78° for the inclination of the Uranian equator, so that the rotation of the planet is likely to be direct. If so, the extraordinary anomaly is here met with of a satellite-system circulating in a direction opposite to that of its primary’s rotation.

Uranus can at times be perceived with the naked eye. Indian traditions of an eighth “dark” planet have been thought to refer to it, and its slow course among the stars had been noted by savage tribes long before Herschel singled it out from them by its tiny disc. It is about three times brighter than Vesta; and Mr. Proctor stated that “in the summer of 1887 they were comparable under favourable conditions,” when both, in the transparent skies of Florida, were “quite conspicuous without telescopic aid.” Twenty chances of discovering Uranus were missed before it came to Herschel’s turn. So many times it had been located or catalogued as a fixed star by astronomers far from indifferent to immortal fame.

Neptune is much nearer to the sun than it ought to be. Both Leverrier and Adams assumed that Bode’s law would hold good for the planet still below the horizon of knowledge; they could do no otherwise; yet the rule played them false. Some have even asserted paradoxically that the planet found was not the planet sought. In point of fact, the distance of the theoretical Neptune is thirty-eight, that of the real Neptune thirty astronomical units. The mean radius of its orbit measures 2,792 million miles. Hence the sun is reduced to ¹⁄₉₀₀th its terrestrial brilliancy, and could be replaced by 687 full moons. “As seen from Neptune,” Professor Young remarks, “the sun would look very much like a large electric arc lamp at a distance of a few feet. It would give about forty-four millions the light of a first-magnitude star.”^[80] Accordingly, Neptune does not circulate by any means in outer darkness. His orbit, although very slightly eccentric, brings him at perihelion fifty millions of miles nearer to the sun than at aphelion. It makes an angle of less than 2° with the ecliptic, and is traversed, at the rate of 3⅓ miles a second, in a period of 165 years.

Neptune, being fainter than the eighth stellar magnitude, is quite inaccessible to unaided vision. But a good telescope at once displays the seeming star in the guise of a small planetary nebula with a diameter of 2″·433. This mean value, reduced to the mean distance of the planet from the sun, was afforded by Barnard’s measures in 1895 with a power of 1,000 on the Lick refractor.^[81] It corresponds to a linear diameter of 32,900 miles. Neptune accordingly, although only 17 times more massive than the earth, is 72 times more bulky, and composed of materials 4·2 times specifically lighter. Gravity at its surface has almost precisely its terrestrial power. The albedo of Neptune, combining Zöllner’s with Müller’s results, is 0·65; and its spectrum appears identical with that of Uranus. It may be inferred that this planet also is too hot to contain water.

Its satellite is believed to be of about the size of the moon; but since it is 12,000 times more distant, it can be distinguished only with the most powerful telescopes as a star of the fourteenth magnitude. The radius of its orbit measures 225,000, that of our moon 238,000 miles; but Neptune’s attendant completes a circuit in 5 days 21 hours; and it is through this rapidity of movement that the large mass of its primary has been learned. It resembles the moon besides in being solitary, so far as can be ascertained by the most diligent researches; and it is beyond doubt that if any companion-bodies exist they are comparatively small or obscure. That they do exist, appears probable on the face of it.

The one Neptunian satellite emphasises the problems set by the Uranian four. These problems are concerned with the origin and early mechanical relations of the solar system. Here, at its utmost verge, we encounter a decided reversal in the direction of systemic motion—a reversal prepared for, as it might seem, by the nearly vertical position of the Uranian plane of satellite-revolution. This diversity is in no sense “accidental,” as some have unwisely asserted, invoking impacts of comets, and such like futile devices, to account for it; it belongs fundamentally to the design of planetary evolution. Laplace’s scheme has no room for it; Faye’s, constructed expressly to include it, requires that Uranus and Neptune, instead of being the first, should have been the latest formed of all the solar train. And their obviously rudimentary condition favours the suggestion. Neptune’s satellite revolves from east to west in a quasi-circular path, inclined to the ecliptic at an angle of 35°; or, putting it otherwise, it revolves from west to east at an angle of 145°.

As the only member of the solar system exempt from perturbations by a third body (the sun being too remote to cause perceptible deflections), it seemed admirably fitted to discharge the functions of a standard celestial clock, greatly needed, but nowhere to be found in our system.^[82] But in 1886 Mr. Marth drew attention to certain divagations of this “ideal time-keeper” resulting from conspicuous changes in the position and plane of its orbit. They were explained almost simultaneously in 1888 by M. Tisserand,^[83] late director of the Paris Observatory, and by Professor Newcomb.^[84] The disturbance, which, in its mode of production, is analogous to the precession of the equinoxes, results from the polar compression of the Neptunian globe combined with a deviation of the satellite’s motion from its equatorial plane. By the action of the protuberant girdle, a slow gyration of the secondary body’s orbital plane is produced, its inclination to the primary’s equator remaining unchanged. Viewed under a different aspect, the same phenomenon may be described as a retrograde movement, in a period of at least five hundred years, of the pole of the satellite’s orbit round the pole of the planet’s equator. The radius of the circle described cannot be less than 20°, implying a flattening of the Neptunian globe of ¹⁄₈₅th, and may easily amount to 30°, with which an ellipticity of ¹⁄₁₁₅ should be associated. But before the centre of this circle—that is, the pole of Neptune’s axial movement—can be satisfactorily located, several centuries must elapse. At present we may affirm with reasonable certainty: first, that the rotation in question is retrograde, like the satellite’s revolution; secondly, basing the inference upon the comparatively slight ellipticity of Neptune’s figure, that it is much slower than the vertiginous spinning of Jupiter, Saturn, and Uranus.

Uranus and Neptune are, as has been said, companion globes. In bulk and density they differ very slightly; their albedoes are virtually the same, their spectra indistinguishable. They seem perfectly alike in chemical and physical constitution, and to be situated at precisely the same stage of development. Both govern retrograde systems. In Uranus the peculiarity appears as if in an incipient form; in Neptune, strongly accentuated.

Viewed from the position of Neptune, all the planets are morning and evening stars. They are tethered to the chariot-wheels of the sun, instead of having the run of the sky. “The four terrestrial planets,” Professor Young writes, “would be hopelessly invisible, unless with powerful telescopes, and by carefully screening off sunlight. Mars would never reach an elongation of three degrees from the sun; the maximum elongation of the earth would be two, and that of Venus about one and a half degrees. Jupiter, attaining an elongation of about ten degrees, would probably be easily seen somewhat as we see Mercury. Saturn and Uranus would be conspicuous, though the latter is the only planet of the whole system that can be better seen from Neptune than it can be from the earth.”^[85]

To a spectator retreating with the velocity of light, all the planetary cortège would in a few hours disappear, and the sun would shine alone. No sign would remain that his office is purely ministerial—that he exists only to enlighten, rule, and vivify the relatively minute globes shred from his mass in the beginning, maintaining by his attractive power the adjusted movements of the complicated piece of mechanism they constitute. The skies perhaps hold millions of his stamp; every solitary star telescopically visible may be the centre of a planetary scheme like our own; or, on the other hand, our own may, quite conceivably, have no counterpart in the wide universe.

CHAPTER XI.
FAMOUS COMETS.

In the fourth year of the 101st Olympiad (373 B.C.), the Greeks were startled by a celestial portent. They did not, at that time, draw fine distinctions, and posterity would have remained ignorant that the terrifying object was a great comet but for the description of it left by Aristotle, who saw it as a boy at Stagira. It was mid-winter when it flared up from due west at sunset, its narrow, definite tail running “like a road through the constellations” over a third of the heavens. Diodorus relates that it cast shadows like the moon, which implies a very unusual, yet not impossible, degree of brightness. The prompt engulfment by an earthquake and its attendant tidal wave of the Achaean towns, Helice and Bura, justified the apprehensions it aroused. It never came back to retrieve its reputation. During at least two thousand subsequent years, such objects lay under the ban of popular superstition; and the counts upon which they were accused of malefic influence were so many and so vague that acquittal was impossible. Their respect of persons was notorious; nor were they consistent in their dealings with the great, to whom alone they paid individual attention. A comet marked the apotheosis of the great Julius; a comet announced the death of Constantine; a comet illuminated the cradle of Napoleon.

The very word “comet” takes us back to the Stagyrite; for it is derived from the Greek word κόμη, hair, and signifies a hirsute star. Shakspeare’s “crystal tresses” represent what we now, in homely fashion, call the “tail,” while the “nucleus” and “coma” make up the “head.” The nucleus, in great comets, shines like a star of the first magnitude, sometimes indeed surpassing the brilliancy of Jupiter. It is usually of measurable dimensions, often of granular texture. The planetary disc, round which the filmy appendages of the comet of December 1618 were displayed, was observed by Cysatus, a Jesuit astronomer at Ingolstadt, to become transformed into the semblance of a star cluster; Hevelius noticed a double nucleus in the comet of 1652; and modern instances of the same kind abound. There is indeed no likelihood that substantial globes are ever included in the construction of comets.

The coma is of immense volume, and extreme tenuity. The rays of faint stars traverse, undimmed and unrefracted, strata of it tens of thousands of miles in thickness. Yet strong lines of structure develop in it through the influence of forces emanating from the sun. As they approach our system out of the depths of space, comets are scarcely distinguishable from round nebulæ, and they relapse into a similar quiescent condition on leaving it. Their temperature must then be very near the absolute zero of cold, since they cannot be supposed either to contain stores of native heat, or to retain stores of borrowed heat. Thus the rapidly augmenting power of solar radiation, as they rush with accelerated velocity nearer and nearer to its source, produce upon them stupendous effects. The nucleus blazes out into a coruscating star; the coma, violently driven off from it, forms multiple envelopes like thin gauze veils, one outside the other, flung round the nucleus on the side next the sun, separated by intervening dark spaces, and diversified by brilliant jets and sectors. The tail is the outcome of a double repulsion. Matter expelled by the nucleus towards the sun is, at a certain point, thrown back to form an immense, oppositely directed appendage, usually convex on the forward side. Some tails resemble hollow cones, being bright at the edges, and dark within: others are traversed by a shining backbone; many, perhaps all, are composite. The magnificent object first seen by Klinkenberg at Haarlem, December 9, 1743, was supplied with six, varying in length from 30° to 44°, each, according to the extant representations, being separately rooted in the head. Grouped into a lustrous fan, they presented a very beautiful and surprising appearance, not again to be displayed until the world and humanity have undergone some unlooked-for changes. For the period of the comet was computed to be one hundred thousand years! Tails, less obviously and splendidly multiplex, are rather the rule than an exception. Or rather, closer observations, chiefly photographic, have made it manifest that the single efflux of nebulous stuff generally designated as a comet’s tail can be analysed into bundles of fibres, into straight rays and curved plumes of light, or into knotted and branching emanations. Homogeneous outflows, such as are seen in drawings, do not really exist. Tails pointing towards the sun have also been occasionally noticed; but they are always feeble. Olbers recorded, however, that, during eight days of January, 1824, the comet then visible had a solar tail of 7°, while its anti-solar tail was only 3½° long.

The great comet of 1680 will always be memorable for having had its orbit calculated by Newton on gravitational principles. It was not unworthy of the distinction. Approaching the sun almost in a straight line, it penetrated the corona at the rate of 370 miles a second, and passing within 140,000 miles of the photosphere, escaped by means of its extraordinary velocity from those perilous precincts. Resulting internal commotions became evident through the rapid development of a tail more than a hundred million miles in length. Newton calculated that particles from the head reached its extremity in two days. He assigned to the comet a highly elliptical orbit traversed in six centuries. But, since its speed might be called parabolic, millenniums may be nearer the mark than centuries. It cannot, therefore, be identified with any earlier apparition.

The comet of 1682 was Halley’s, the predicted return of which, in 1759, was unprecedented and memorable. At its apparition in 1835, valuable observations of a physical kind were made upon it by Bessel at Königsberg, and by Sir John Herschel at the Cape. They were facilitated by the circumstance that this far-travelling body, the perihelion distance of which is 55 million miles, and the aphelion-distance 2½ times that of Neptune, approached the earth on this occasion within 4½ million miles. It was remarkable for singular and sudden changes of aspect. To Bessel the nucleus seemed like a burning rocket. Divergent flames issued from it towards the sun, and he took especial note of a blazing “sector,” which swung like a pendulum to and fro, in a period of 4⅗ days. These emanations, accumulating at the surface where the solar balanced the cometary repulsive force, were then swept back, as if by a tempestuous wind, to form a tail, which, on October 15, measured at least 24°. The conviction was forced upon him that the body in which these wonderful processes were going on was affected by opposite polarities; and he fully concurred with Olbers in the opinion that tail-production was a purely electrical phenomenon.

During some time before and after its perihelion passage on November 16, the comet wore the disguise of a star. All its hairy appendages had vanished. On the 23rd of January, 1836, it was sharply stellar; twenty-four hours later it had acquired, besides a twenty-fold increase of light, a disc like that of the planet Neptune, enclosed in a nebulous sheath of about fourfold breadth. Later in its career, Sir John Herschel^[86] observed the nucleus under the form of “a miniature comet, having a nucleus, head, and tail of its own, perfectly distinct, and considerably exceeding in intensity of light the nebulous disc or envelope” containing it, which was, properly speaking, the “head” of the comet. At last, on May 5, through the progress of distension, the last thin shred of its substance melted into the sky. The next return of Halley’s comet, somewhat accelerated by Jupiter’s influence, is looked for in the year 1910.

The “vintage comet” lingered in northern skies during 510 days—from March 26, 1811, until August 17, 1812. It was attentively observed by Sir William Herschel, who gathered from it the then new truth that comets are self-luminous bodies. “The quality of giving out light,” he acutely remarked, “is immensely increased by an approach to the sun.” But he failed to persuade his contemporaries or successors. His inference had to wait for spectroscopic demonstration. The nucleus of the comet of 1811 he found to measure 428 miles. It showed a ruddy hue, and was eccentrically placed within a greenish-blue “planetary body” 127,000 miles in diameter. This was again enclosed in a shining atmosphere about four times as wide, round which was flung an envelope of a yellow tint, forming a thin hemispherical shell on the side next the sun, and continued indefinitely away from the sun as the hollow cone of the tail. Owing to this mode of construction, the space between the head and the hemispherical sheath, as well as the central part of the tail, appeared dark. The latter extended, in October, over 100 million miles of space, and was 15 million miles broad. Its soft radiance resembled that of the Milky Way, side by side with which it ran on November 9, 1811. The comet’s path lay entirely outside the earth’s orbit, and Argelander assigned to it a period of 3,065 years. The restriction was needless. Between a period of infinite length, and one of 3,000, or 1,000 years, no valid distinction can, where comets are in question, be drawn. The short sections of their tracks observable from the earth might belong equally well to parabolas or to the far-stretching ellipses which such protracted periods imply.

The apparition of 1811 suggested to Olbers the “electrical theory” of comets’ tails. The uncommon impressiveness with which it displayed not uncommon phenomena, was perhaps a result of its considerable distance from the sun, owing to which the interior force obtained an advantage over the exterior, and the locus of equilibrium between solar and cometary repulsion was pushed back further than usual from the nucleus.^[87] He calculated that the materials of the tail spent 11 minutes in the journey from its root to its tip, indicating ejection by a force greatly more powerful than the opposing force of gravity. Olbers anticipated the modern view that chemical differences determine the shapes of comets’ tails, the various species of matter being diversely acted upon by electrical repulsion. The long, straight ray, for instance, issuing from the comet of 1807, must, he perceived, have been composed of particles much more energetically repelled than those aggregated in the inflected plume with which it was associated. The curvature of these appendages, in fact, depends upon the relation between the orbital velocity of the comet and the velocity of ejection imparted to their constituent molecules. It has to be borne in mind, however, that while curved tails may appear straight in projection, straight tails can never appear curved

Olbers’ classification of comets is still of great significance. He divided them into: