RECENT COMETS

On the 2nd of June, 1858, Giambattista Donati discovered at Florence a feeble round nebulosity in the constellation Leo, about one-tenth the diameter of the full moon. It proved to be a comet approaching the sun. But it changed little in apparent place or brightness for some weeks. The gradual development of a central condensation of light was the first symptom of coming splendour. At Harvard, in the middle of July, a strong stellar nucleus was seen; on August 14 a tail began to be thrown out. As the comet wanted still over six weeks of the time of its perihelion-passage, it was obvious that great things might be expected of it. They did not fail of realisation.

Not before the early days of September was it generally recognised with the naked eye, though it had been detected without a glass at Pulkowa, August 19. But its growth was thenceforward surprisingly rapid, as it swept with accelerated motion under the hindmost foot of the Great Bear, and past the starry locks of Berenice. A sudden leap upward in lustre was noticed on September 12, when the nucleus shone with about the brightness of the pole-star, and the tail, notwithstanding large foreshortening, could be traced with the lowest telescopic power over six degrees of the sphere. The appendage, however, attained its full development only after perihelion, September 30, by which time, too, it lay nearly square to the line of sight from the earth. On October 10 it stretched in a magnificent scimitar-like curve over a third and upwards of the visible hemisphere, representing a real extension in space of fifty-four million miles. But the most striking view was presented on October 5, when the brilliant star Arcturus became involved in the brightest part of the tail, and during many hours contributed, its lustre undiminished by the interposed nebulous screen, to heighten the grandeur of the most majestic celestial object of which living memories retain the impress. Donati's comet was, according to Admiral Smyth's testimony,[1188] outdone "as a mere sight-object" by the great comet of 1811; but what it lacked in splendour, it surely made up in grace, and variety of what we may call "scenic" effects.

Some of these were no less interesting to the student than impressive to the spectator. At Pulkowa, on the 16th September, Winnecke,[1189] the first director of the Strasburg Observatory, observed a faint outer envelope resembling a veil of almost evanescent texture flung somewhat widely over the head. Next evening, the first of the "secondary" tails appeared, possibly as part of the same phenomenon. This was a narrow straight ray, forming a tangent to the strong curve of the primary tail, and reaching to a still greater distance from the nucleus. It continued faintly visible for about three weeks, during part of which time it was seen in duplicate. For from the chief train itself, at a point where its curvature abruptly changed, issued, as if through the rejection of some of its materials, a second beam nearly parallel to the first, the rigid line of which contrasted singularly with the softly diffused and waving aspect of the plume of light from which it sprang. Olbers's theory of unequal repulsive forces was never more beautifully illustrated. The triple tail seemed a visible solar analysis of cometary matter.

The processes of luminous emanation going on in this body forcibly recalled the observations made on the comets of 1744 and 1835. From the middle of September, the nucleus, estimated by Bond to be under five hundred miles in diameter, was the centre of action of the most energetic kind. Seven distinct "envelopes" were detached in succession from the nebulosity surrounding the head, and after rising towards the sun during periods of from four to seven days, finally cast their material backward to form the right and left branches of the great train. The separation of these by an obscure axis—apparently as black, quite close up to the nucleus, as the sky—indicated for the tail a hollow, cone-like structure;[1190] while the repetition of certain spots and rays in the same corresponding situation on one envelope after another served to show that the nucleus—to some local peculiarity of which they were doubtless due—had no proper rotation, but merely shifted sufficiently on an axis to preserve the same aspect towards the sun as it moved round it.[1191] This observation of Bond's was strongly confirmatory of Bessel's hypothesis of opposite polarities in such bodies' opposite sides.

The protrusion towards the sun, on September 25, of a brilliant luminous fan-shaped sector completed the resemblance to Halley's comet. The appearance of the head was now somewhat that of a "bat's-wing" gaslight. There were, however, no oscillations to and fro, such as Bessel had seen and speculated upon in 1835. As the size of the nucleus contracted with approach to perihelion, its intensity augmented. On October 2, it outshone Arcturus, and for a week or ten days was a conspicuous object half an hour after sunset. Its lustre—setting aside the light derived from the tail—was, at that date, 6,300 times what it had been on June 15, though theoretically—taking into account, that is, only the differences of distance from sun and earth—it should have been only 1/33 of that amount. Here, it might be thought, was convincing evidence of the comet itself becoming ignited under the growing intensity of the solar radiations. Yet experiments with the polariscope were interpreted in an adverse sense, and Bond's conclusion that the comet sent us virtually unmixed reflected sunshine was generally acquiesced in. It was, nevertheless, negatived by the first application of the spectroscope to these bodies.

Very few comets have been so well or so long observed as Donati's. It was visible to the naked eye during 112 days; it was telescopically discernible for 275, the last observation having been made by Mr. William Mann at the Cape of Good Hope, March 4, 1859. Its course through the heavens combined singularly with the orbital place of the earth to favour curious inspection. The tail, when near its greatest development, lost next to nothing by the effects of perspective, and at the same time lay in a plane sufficiently inclined to the line of sight to enable it to display its exquisite curves to the greatest advantage. Even the weather was, on both sides of the Atlantic, propitious during the period of greatest interest, and the moon as little troublesome as possible. The volume compiled by the younger Bond is a monument to the care and skill with which these advantages were turned to account. Yet this stately apparition marked no turning-point in the history of cometary science. By its study knowledge was indeed materially advanced, but along the old lines. No quick and vivid illumination broke upon its path. Quite insignificant objects—as we have already partly seen—have often proved more vitally instructive.

Donati's comet has been identified with no other. Its path is an immensely elongated ellipse, lying in a plane far apart from that of the planetary movements, carrying it at perihelion considerably within the orbit of Venus, and at aphelion out into space to 5-1/2 times the distance from the sun of Neptune. The entire circuit occupies over 2,000 years, and is performed in a retrograde direction, or against the order of the Signs. Before its next return, about the year 4000 A.D., the enigma of its presence and its purpose may have been to some extent—though we may be sure not completely—penetrated.

On June 30, 1861, the earth passed, for the second time in the century, through the tail of a great comet. Some of our readers may remember the unexpected disclosure, on the withdrawal of the sun below the horizon on that evening, of an object so remarkable as to challenge universal attention. A golden-yellow planetary disc, wrapt in dense nebulosity, shone out while the June twilight of these latitudes was still in its first strength. The number and complexity of the envelopes surrounding the head produced, according to the late Mr. Webb,[1192] a magnificent effect. Portions of six distinct emanations were traceable. "It was as though a number of light, hazy clouds were floating round a miniature full moon." As the sky darkened the tail emerged to view.[1193] Although in brightness and sharpness of definition it could not compete with the display of 1858, its dimensions proved to be extraordinary. It reached upwards beyond the zenith when the head had already set. By some authorities its extreme length was stated at 118°, and it showed no trace of curvature. Most remarkable, however, was the appearance of two widely divergent rays, each pointing towards the head, though cut off from it by sky-illumination, of which one was seen by Mr. Webb, and both by Mr. Williams at Liverpool, a quarter of an hour before midnight. There seems no doubt that Webb's interpretation was the true one, and that these beams were, in fact, "the perspective representation of a conical or cylindrical tail, hanging closely above our heads, and probably just being lifted up out of our atmosphere."[1194] The cometary train was then rapidly receding from the earth, so that the sides of the "outspread fan" of light shown by it when we were right in the line of its axis must have appeared (as they did) to close up in departure. The swiftness with which the visually opened fan shut proved its vicinity; and, indeed, Mr. Hind's calculations showed that we were not so much near as actually within its folds at that very time.

Already M. Liais, from his observations at Rio de Janeiro, June 11 to 14, and Mr. Tebbutt, by whom the comet was discovered in New South Wales on May 13, had anticipated such an encounter, while the former subsequently proved that it must have occurred in such a way as to cause an immersion of the earth in cometary matter to a depth of 300,000 miles.[1195] The comet then lay between the earth and the sun at a distance of about fourteen million miles from the former; its tail stretched outward just along the line of intersection of its own with the terrestrial orbit to an extent of fifteen million miles; so that our globe, happening to pass at the time, found itself during some hours involved in the flimsy appendage.

No perceptible effects were produced by the meeting; it was known to have occurred by theory alone. A peculiar glare in the sky, thought by some to have distinguished the evening of June 30, was, at best, inconspicuous. Nor were there any symptoms of unusual electric excitement. The Greenwich instruments were, indeed, disturbed on the following night, but it would be rash to infer that the comet had art or part in their agitation.

The perihelion-passage of this body occurred June 11, 1861; and its orbit has been shown by M. Kreutz of Bonn, from a very complete investigation founded on observations extending over nearly a year, to be an ellipse traversed in a period of 409 years.[1196]

Towards the end of August, 1862, a comet became visible to the naked eye high up in the northern hemisphere, with a nucleus equalling in brightness the lesser stars of the Plough and a feeble tail 20° in length. It thus occupied quite a secondary position among the members of its class. It was, nevertheless, a splendid object in comparison with a telescopic nebulosity discovered by Tempel at Marseilles, December 19, 1865. This, the sole comet of 1866, slipped past perihelion, January 11, without pomp of train or other appendages, and might have seemed hardly worth the trouble of pursuing. Fortunately, this was not the view entertained by observers and computers; since upon the knowledge acquired of the movements of these two bodies has been founded one of the most significant discoveries of modern times. The first of them is now styled the comet (1862 iii.) of the August meteors, the second (1866 i.) that of the November meteors. The steps by which this curious connection came to be ascertained were many, and were taken in succession by a number of individuals. But the final result was reached by Schiaparelli of Milan, and remains deservedly associated with his name.

The idea prevalent in the eighteenth century as to the nature of shooting stars was that they were mere aerial ignes fatui—inflammable vapours accidentally kindled in our atmosphere. But Halley had already entertained the opinion of their cosmical origin; and Chladni in 1794 formally broached the theory that space is filled with minute circulating atoms, which, drawn by the earth's attraction, and ignited by friction in its gaseous envelope, produce the luminous effects so frequently witnessed.[1197] Acting on his suggestion, Brandes and Benzenberg, two students at the University of GÖttingen, began in 1798 to determine the heights of falling stars by simultaneous observations at a distance. They soon found that they move with planetary velocities in the most elevated regions of our atmosphere, and by the ascertainment of this fact laid a foundation of distinct knowledge regarding them. Some of the data collected, however, served only to perplex opinion, and even caused Chladni temporarily to renounce his. Many high authorities, headed by Laplace in 1802, declared for the lunar-volcanic origin of meteorites; but thought on the subject was turbid, and inquiry seemed only to stir up the mud of ignorance. It needed one of those amazing spectacles, at which man assists, no longer in abject terror for his own frail fortunes, but with keen curiosity and the vivid expectation of new knowledge, to bring about a clarification.

On the night of November 12-13, 1833, a tempest of falling stars broke over the earth. North America bore the brunt of its pelting. From the Gulf of Mexico to Halifax, until daylight with some difficulty put an end to the display, the sky was scored in every direction with shining tracks and illuminated with majestic fireballs. At Boston the frequency of meteors was estimated to be about half that of flakes of snow in an average snowstorm. Their numbers, while the first fury of their coming lasted, were quite beyond counting; but as it waned, a reckoning was attempted, from which it was computed, on the basis of that much diminished rate, that 240,000 must have been visible during the nine hours they continued to fall.[1198]

Now there was one very remarkable feature common to the innumerable small bodies which traversed, or were consumed in our atmosphere that night. They all seemed to come from the same part of the sky. Traced backward, their paths were invariably found to converge to a point in the constellation Leo. Moreover, that point travelled with the stars in their nightly round. In other words, it was entirely independent of the earth and its rotation. It was a point in inter-planetary space.

The effective perception of this fact[1199] amounted to a discovery, as Olmsted and Twining, who had "simultaneous ideas" on the subject, were the first to realize. Denison Olmsted was then Professor of Mathematics in Yale College. He showed early in 1834[1200] that the emanation of the showering meteors from a fixed "radiant" proved their approach to the earth along nearly parallel lines, appearing to diverge by an effect of perspective; and that those parallel lines must be sections of orbits described by them round the sun and intersecting that of the earth. For the November phenomenon was now seen to be a periodical one. On the same night of the year 1832, although with less dazzling and universal splendour than in America in 1833, it had been witnessed over great part of Europe and in Arabia. Olmsted accordingly assigned to the cloud of cosmical particles (or "comet," as he chose to call it), by terrestrial encounters with which he supposed the appearances in question to be produced, a period of about 182 days; its path a narrow ellipse, meeting, near its farthest end from the sun, the place occupied by the earth on November 12.

Once for all, then, as the result of the star-fall of 1833, the study of luminous meteors became an integral part of astronomy. Their membership of the solar system was no longer a theory or a conjecture—it was an established fact. The discovery might be compared to, if it did not transcend in importance, that of the asteroidal group. "C'est un nouveau monde planÉtaire," Arago wrote,[1201] "qui commence À se rÉvÉler À nous."

Evidences of periodicity continued to accumulate. It was remembered that Humboldt and Bonpland had been the spectators at Cumana, after midnight on November 12, 1799, of a fiery shower little inferior to that of 1833, and reported to have been visible from the equator to Greenland. Moreover, in 1834 and some subsequent years, there were waning repetitions of the display, as if through the gradual thinning-out of the meteoric supply. The extreme irregularity of its distribution was noted by Olbers in 1837, who conjectured that we might have to wait until 1867 to see the phenomenon renewed on its former scale of magnificence.[1202] This was the first hint of a thirty-three or thirty-four year period.

The falling stars of November did not alone attract the attention of the learned. Similar appearances were traditionally associated with August 10 by the popular phrase in which they figured as "the tears of St. Lawrence." But the association could not be taken on trust from mediÆval authority. It had to be proved scientifically, and this Quetelet of Brussels succeeded in doing in December, 1836.[1203]

A second meteoric revolving system was thus shown to exist. But its establishment was at once perceived to be fatal to the "cosmical cloud" hypothesis of Olmsted. For if it be a violation of probability to attribute to one such agglomeration a period of an exact year, or sub-multiple of a year, it would be plainly absurd to suppose the movements of two or more regulated by such highly artificial conditions. An alternative was proposed by Adolf Erman of Berlin in 1839.[1204] No longer in clouds, but in closed rings, he supposed meteoric matter to revolve round the sun. Thus the mere circumstance of intersection by a meteoric of the terrestrial orbit, without any coincidence of period, would account for the earth meeting some members of the system at each annual passage through the "node" or point of intersection. This was an important step in advance, yet it decided nothing as to the forms of the orbits of such annular assemblages; nor was it followed up in any direction for a quarter of a century.

Hubert A. Newton took up, in 1864,[1205] the dropped thread of inquiry. The son of a mathematical mother, he attained, at the age of twenty-five, to the dignity of Professor of Mathematics in Yale University, and occupied the post until his death in 1896. The diversion of his powers, however, from purely abstract studies stimulated their effective exercise, and constituted him one of the founders of meteoric astronomy.

A search through old records carried the November phenomenon back to the year 902 A.D., long distinguished as "the year of the stars." For in the same night in which Taormina was captured by the Saracens, and the cruel Aghlabite tyrant Ibrahim ibn Ahmed died "by the judgment of God" before Cosenza, stars fell from heaven in such abundance as to amaze and terrify beholders far and near. This was on October 13, and recurrences were traced down through the subsequent centuries, always with a day's delay in about seventy years. It was easy, too, to derive from the dates a cycle of 33-1/4 years, so that Professor Newton did not hesitate to predict the exhibition of an unusually striking meteoric spectacle on November 13-14, 1866.[1206]

For the astronomical explanation of the phenomena, recourse was had to a method introduced by Erman of computing meteoric orbits. It was found, however, that conspicuous recurrences every thirty-three or thirty-four years could be explained on the supposition of five widely different periods, combined with varying degrees of extension in the revolving group. Professor Newton himself gave the preference to the shortest—of 354-1/2 days, but indicated the means of deciding with certainty upon the true one. It was furnished by the advancing motion of the node, or that day's delay of the November shower every seventy years, which the old chronicles had supplied data for detecting. For this is a strictly measurable effect of gravitational disturbance by the various planets, the amount of which naturally depends upon the course pursued by the disturbed bodies. Here the great mathematical resources of Professor Adams were brought to bear. By laborious processes of calculation, he ascertained that four out of Newton's five possible periods were entirely incompatible with the observed nodal displacement, while for the fifth—that of 33-1/4 years—a perfectly harmonious result was obtained.[1207] This was the last link in the chain of evidence proving that the November meteors—or "Leonids," as they had by that time come to be called—revolve round the sun in a period of 33·27 years, in an ellipse spanning the vast gulf between the orbits of the earth and Uranus, the group being so extended as to occupy nearly three years in defiling past the scene of terrestrial encounters. But before it was completed in March, 1867, the subject had assumed a new aspect and importance.

Professor Newton's prediction of a remarkable star-shower in November, 1866, was punctually fulfilled. This time, Europe served as the main target of the celestial projectiles, and observers were numerous and forewarned. The display, although, according to Mr. Baxendell's memory,[1208] inferior to that of 1833, was of extraordinary impressiveness. Dense crowds of meteors, equal in lustre to the brightest stars, and some rivalling Venus at her best,[1209] darted from east to west across the sky with enormous apparent velocities, and with a certain determinateness of aim, as if let fly with a purpose, and at some definite object.[1210] Nearly all left behind them trains of emerald green or clear blue light, which occasionally lasted many minutes, before they shrivelled and curled up out of sight. The maximum rush occurred a little after one o'clock on the morning of November 14, when attempts to count were overpowered by frequency. But during a previous interval of seven minutes five seconds, four observers at Mr. Bishop's observatory at Twickenham reckoned 514, and during an hour 1,120.[1211] Before daylight the earth had fairly cut her way through the star-bearing stratum; the "ethereal rockets" had ceased to fly.

This event brought the subject of shooting stars once more vividly to the notice of astronomers. Schiaparelli had, indeed, been already attracted by it. The results of his studies were made known in four remarkable letters, addressed, before the close of the year 1866, to Father Secchi, and published in the Bulletino of the Roman Observatory.[1212] Their upshot was to show, in the first place, that meteors possess a real velocity considerably greater than that of the earth, and travel, accordingly, to enormously greater distances from the sun along tracks resembling those of comets in being very eccentric, in lying at all levels indifferently, and in being pursued in either direction. It was next inferred that comets and meteors equally have an origin foreign to the solar system, but are drawn into it temporarily by the sun's attraction, and occasionally fixed in it by the backward pull of some planet. But the crowning fact was reserved for the last. It was the astonishing one that the August meteors move in the same orbit with the bright comet of 1862—that the comet, in fact, is but a larger member of the family named "Perseids" because their radiant point is situated in the constellation Perseus.

This discovery was quickly capped by others of the same kind. Leverrier published, January 21, 1867,[1213] elements for the November swarm, founded on the most recent and authentic observations; at once identified by Dr. C. F. W. Peters of Altona with Oppolzer's elements for Tempel's comet of 1866.[1214] A few days later, Schiaparelli, having recalculated the orbit of the meteors from improved data, arrived at the same conclusion; while Professor Weiss of Vienna pointed to the agreement between the orbits of a comet which had appeared in 1861 and of a star-shower found to recur on April 20 (LyraÏds), as well as between those of Biela's comet and certain conspicuous meteors of November 28.[1215]

These instances do not seem to be exceptional. The number of known or suspected accordances of cometary tracks with meteor streams contained in a list drawn up in 1878[1216] by Professor Alexander S. Herschel (who has made the subject peculiarly his own) amounts to seventy-six; although the four first detected still remain the most conspicuous, and perhaps the only absolutely sure examples of a relation as significant as it was, to most astronomers, unexpected.

There had, indeed, been anticipatory ideas. Not that Kepler's comparison of shooting stars to "minute comets," or Maskelyne's "forse risulterÀ che essi sono comete," in a letter to the Abate Cesaris, December 12, 1774,[1217] need count for much. But Chladni, in 1819,[1218] considered both to be fragments or particles of the same primitive matter, irregularly scattered through space as nebulÆ; and Morstadt of Prague suggested about 1837[1219] that the meteors of November might be dispersed atoms from the tail of Biela's comet, the path of which is cut across by the earth near that epoch. Professor Kirkwood, however, by a luminous intuition, penetrated the whole secret, so far as it has yet been made known. In an article published, or rather buried, in the Danville Quarterly Review for December, 1861, he argued, from the observed division of Biela, and other less noted instances of the same kind, that the sun exercises a "divellent influence" on the nuclei of comets, which may be presumed to continue its action until their corporate existence (so to speak) ends in complete pulverisation. "May not," he continued, "our periodic meteors be the dÉbris of ancient but now disintegrated comets, whose matter has become distributed round their orbits?"[1220]

The gist of Schiaparelli's discovery could not be more clearly conveyed. For it must be borne in mind that with the ultimate destiny of comets' tails this had nothing to do. The tenuous matter composing them is, no doubt, permanently lost to the body from which it emanated; but science does not pretend to track its further wanderings through space. It can, however, state categorically that these will no longer be conducted along the paths forsaken under solar compulsion. From the central, and probably solid parts of comets, on the other hand, are derived the granules by the swift passage of which our skies are seamed with periodic fires. It is certain that a loosely agglomerated mass (such as cometary nuclei most likely are) must gradually separate through the unequal action of gravity on its various parts—through, in short, solar tidal influence. Thenceforward its fragments will revolve independently in parallel orbits, at first as a swarm, finally—when time has been given for the full effects of the lagging of the slower moving particles to develop—as a closed ring. The first condition is still, more or less, that of the November meteors; those of August have already arrived at the second. For this reason, Leverrier pronounced, in 1867, the Perseid to be of older formation than the Leonid system. He even assigned a date at which the introduction of the last-named bodies into their present orbit was probably effected through the influence of Uranus. In 126 A.D. a close approach must have taken place between the planet and the parent comet of the November stars, after which its regular returns to perihelion, and the consequent process of its disintegration, set in. Though not complete, it is already far advanced.

The view that meteorites are the dust of decaying comets was now to be put to a definite test of prediction. Biela's comet had not been seen since its duplicate return in 1852. Yet it had been carefully watched for with the best telescopes; its path was accurately known; every perturbation it could suffer was scrupulously taken into account. Under these circumstances, its repeated failure to come up to time might fairly be thought to imply a cessation from visible existence. Might it not, however, be possible that it would appear under another form—that a star-shower might have sprung from and would commemorate its dissolution?

An unusually large number of falling stars were seen by Brandes, December 6, 1798. Similar displays were noticed in the years 1830, 1838, and 1847, and the point from which they emanated was shown by Heis at Aix-la-Chapelle to be situated near the bright star ? AndromedÆ.[1221] Now this is precisely the direction in which the orbit of Biela's comet would seem to lie, as it runs down to cut the terrestrial track very near the place of the earth at the above dates. The inference was, then, an easy one, that the meteors were pursuing the same path with the comet; and it was separately arrived at, early in 1867, by Weiss, D'Arrest, and Galle.[1222] But Biela travels in the opposite direction to Tempel's comet and its attendant "Leonids"; its motion is direct, or from west to east, while theirs is retrograde. Consequently, the motion of its node is in the opposite direction too. In other words, the meeting-place of its orbit with that of the earth retreats (and very rapidly) along the ecliptic instead of advancing. So that if the "Andromedes" stood in the supposed intimate relation to Biela's comet, they might be expected to anticipate the times of their recurrence by as much as a week in half a century. All doubt as to the fact may be said to have been removed by Signor Zezioli's observation of the annual shower in more than usual abundance at Bergamo, November 30, 1867.

The missing comet was next due at perihelion in the year 1872, and the probability was contemplated by both Weiss and Galle of its being replaced by a copious discharge of falling stars. The precise date of the occurrence was not easily determinable, but Galle thought the chances in favour of November 28. The event anticipated the prediction by twenty-four hours. Scarcely had the sun set in Western Europe on November 27, when it became evident that Biela's comet was shedding over us the pulverised products of its disintegration. The meteors came in volleys from the foot of the Chained Lady, their numbers at times baffling the attempt to keep a reckoning. At Moncalieri, about 8 p.m., they constituted (as Father Denza said[1223]) a "real rain of fire." Four observers counted, on an average, four hundred each minute and a half; and not a few fireballs, equalling the moon in diameter, traversed the sky. On the whole, however, the stars of 1872, though about equally numerous, were less brilliant than those of 1866; the phosphorescent tracks marking their passage were comparatively evanescent and their movements sluggish. This is easily understood when we remember that the Andromedes overtake the earth, while the Leonids rush to meet it; the velocity of encounter for the first class of bodies being under twelve, for the second above forty-four miles a second. The spectacle was, nevertheless, magnificent. It presented itself successively to various parts of the earth, from Bombay and the Mauritius to New Brunswick and Venezuela, and was most diligently and extensively observed. Here it had well-nigh terminated by midnight.[1224]

It was attended by a slight aurora, and although Tacchini had telegraphed that the state of the sun rendered some show of polar lights probable, it has too often figured as an accompaniment of star-showers to permit the coincidence to rank as fortuitous. Admiral Wrangel was accustomed to describe how, during the prevalence of an aurora on the Siberian coast, the passage of a meteor never failed to extend the luminosity to parts of the sky previously dark;[1225] and an enhancement of electrical disturbance may well be associated with the flittings of such cosmical atoms.

A singular incident connected with the meteors of 1872 has now to be recounted. The late Professor Klinkerfues, who had observed them very completely at GÖttingen, was led to believe that not merely the dÉbris strewn along its path, but the comet itself must have been in immediate proximity to the earth during their appearance.[1226] If so, it might be possible, he thought, to descry it as it retreated in the diametrically opposite direction from that in which it had approached. On November 30, accordingly, he telegraphed to Mr. Pogson, the Madras astronomer, "Biela touched earth November 27; search near Theta Centauri"—the "anti-radiant," as it is called, being situated close to that star. Bad weather prohibited observation during thirty-six hours, but when the rain clouds broke on the morning of December 2, there a comet was, just in the indicated position. In appearance it might have passed well enough for one of the Biela twins. It had no tail, but a decided nucleus, and was about 45 seconds across, being thus altogether below the range of naked-eye discernment. It was again observed December 3, when a short tail was perceptible; but overcast skies supervened, and it has never since been seen. Its identity accordingly remains in doubt. It seems tolerably certain, however, that it was not the lost comet, which ought to have passed that spot twelve weeks earlier, and was subject to no conceivable disturbance capable of delaying to that extent its revolution. On the other hand, there is the strongest likelihood that it belonged to the same system[1227]—that it was a third fragment, torn from the parent-body of the Andromedes at a period anterior to our first observations of it.

In thirteen years Biela's comet (or its relics) travels nearly twice round its orbit, so that a renewal of the meteoric shower of 1872 was looked for on the same day of the year 1885, the probability being emphasised by an admonitory circular from Dunecht. Astronomers were accordingly on the alert, and were not disappointed. In England, observation was partially impeded by clouds; but at Malta, Palermo, Beyrout, and other southern stations, the scene was most striking. The meteors were both larger and more numerous than in 1872. Their numbers in the densest part of the drift were estimated by Professor Newton at 75,000 per hour, visible from one spot to so large a group of spectators that practically none could be missed. Yet each of these multitudinous little bodies was found by him to travel in a clear cubical space of which the edge measured twenty miles![1228] Thus the dazzling effect of a luminous throng was produced without jostling or overcrowding, by particles, it might almost be said, isolated in the void.

Their aspect was strongly characteristic of the Andromede family of meteors. "They invariably," Mr. Denning wrote,[1229] "traversed short paths with very slow motions, and became extinct in evolved streams of yellowish sparks." The conclusion seemed obvious "that these meteors are formed of very soft materials, which expand while incalescent, and are immediately crumbled and dissipated into exiguous dust."

The Biela meteors of 1885 did not merely gratify astronomers with a fulfilled prediction, but were the means of communicating to them some valuable information. Although their main body was cut through by the moving earth in six hours, and was not more than 100,000 miles across, skirmishers were thrown out to nearly a million miles on either side of the compact central battalions. Members of the system were, on the 26th of November, recorded by Mr. Denning at the hourly rate of about 130; and they did not wholly cease to be visible until December 1. They afforded besides a particularly well-marked example of that diffuseness of radiation previously observed in some less conspicuous displays. Their paths seemed to diverge from an area rather than from a point in the sky. They came so ill to focus that divergences of several degrees were found between the most authentically determined radiants. These incongruities are attributed by Professor Newton to the irregular shape of the meteoroids producing unsymmetrical resistance from the air, and hence causing them to glance from their original direction on entering it. Thus, their luminous tracks did not always represent (even apart from the effects of the earth's attraction) the true prolongation of their course through space.

The Andromedes of 1872 were laggards behind the comet from which they sprang; those of 1885 were its avant-couriers. That wasted and disrupted body was not due at the node until January 26, 1886, sixty days, that is, after the earth's encounter with its meteoric fragments. These are now probably scattered over more than five hundred million miles of its orbits;[1230] yet Professor Newton considers that all must have formed one compact group with Biela at the time of its close approach to Jupiter about the middle of 1841. For otherwise both comet and meteorites could not have experienced, as they seem to have done, the same kind and amount of disturbance. The rapidity of cometary disintegration is thus curiously illustrated.

A short-lived persuasion that the missing heavenly body itself had been recovered, was created by Mr. Edwin Holms's discovery, at London, November 6, 1892, of a tolerably bright, tailless comet, just in a spot which Biela's comet must have traversed in approaching the intersection of its orbit with that of the earth. A hasty calculation by Berberich assigned elements to the newcomer seeming not only to ratify the identity, but to promise a quasi-encounter with the earth on November 21. The only effect of the prediction, however, was to raise a panic among the negroes of the Southern States of America. The comet quietly ignored it, and moved away from instead of towards the appointed meeting-place. Its projection, then, on the night of its discovery, upon a point of the Biela-orbit was by a mere caprice of chance. North America, nevertheless, was visited on November 23 by a genuine Andromede shower. Although the meteors were less numerous than in 1885, Professor Young estimated that 30,000, at the least, of their orange fire-streaks came, during five hours, within the range of view at Princeton.[1231] BrÉdikhine estimated the width of the space containing them at about 2,700,000 miles.[1232] The anticipation of their due time by four days implied—if they were a prolongation of the main Biela group, the nucleus of which passed the spot of encounter five months previously—a recession of the node since 1885 by no less than three degrees. Unless, indeed, Mr. Denning were right in supposing the display to have proceeded from "an associated branch of the main swarm through which we passed in 1872 and 1885."[1233] The existence of separated detachments of Biela meteors, due to disturbing planetary action, was contemplated as highly probable by Schiaparelli.[1234] Such may have been the belated flights met with in 1830, 1838, 1841, and 1847, and such the advance flight plunged through in 1892. A shower looked for November 23, 1899, did not fall, and no further display from this quarter is probable until November 17, 1905, although one is possible a year earlier.[1235]

The Leonids, through the adverse influence of Jupiter and Saturn, inflicted upon multitudes of eager watchers a still more poignant disappointment. A dense part of the swarm, having nearly completed a revolution since 1866, should, travelling normally, have met the earth November 15, 1899; in point of fact, it swerved sunward, and the millions of meteorites which would otherwise have been sacrificed for the illumination of our skies escaped a fiery doom. The contingency had been forecast in the able calculations of Dr. Johnstone Stoney and Dr. A. M. W. Downing,[1236] superintendent of the Nautical Almanac Office; but the verification scarcely compensated the failure. Nor was the situation retrieved in the following years. Only ragged fringes of the great tempest-cloud here and there touched our globe. As the same investigators warned us to expect, the course of the meteorites had been not only rendered sinuous by perturbation, but also broken and irregular. We can no longer count upon the Leonids. Their glory, for scenic purposes, is departed. The comet associated with them also evaded observation. Although it doubtless kept its tryst with the sun in the spring of 1899, the attendant circumstances were too unfavourable to allow it to be seen from the earth.[1237] By an almost fantastic coincidence, nevertheless, a faint comet was photographed, November 14, 1898,[1238] by Dr. Chase, of the Yale College Observatory, close to the Leonid radiant, whither a "meteorograph" was directed with a view to recording trails left by precursors of the main Leonid body. A promising start, too, was made on the same occasion with meteoric researches from sensitive plates.[1239] Indeed, Schaeberle and Colton[1240] had already, in 1896, determined the height of a Leonid by means of photographs taken at stations on different ridges of Mount Hamilton; and Professor Pickering has prosecuted similar work at Harvard, with encouraging results. Everything in this branch of science depends upon how far they can be carried. Without the meteorograph, rigid accuracy in the observation of shooting stars is unattainable, and rigid accuracy is the sine qu non for obtaining exact knowledge.

Biela does not offer the only example of cometary disruption. Setting aside the unauthentic reports of early chroniclers, we meet the "double comet" discovered by Liais at Olinda (Brazil), February 27, 1860, of which the division appeared recent, and about to be carried farther.[1241] But a division once established, separation must continually progress. The periodic times of the fragments will never be identical; one must drop a little behind the other at each revolution, until at length they come to travel in remote parts of nearly the same orbit. Thus the comet predicted by Klinkerfues and discovered by Pogson had already lagged to the extent of twelve weeks, and we shall meet instances farther on where the retardation is counted, not by weeks, but by years. Here original identity emerges only from calculation and comparison of orbits.

Comets, then, die, as Kepler wrote long ago, sicut bombyces filo fundendo. This certainty, anticipated by Kirkwood in 1861, we have at least acquired from the discovery of their generative connection with meteors. Nay, their actual materials become, in smaller or larger proportions, incorporated with our globe. It is not, indeed, universally admitted that the ponderous masses of which, according to DaubrÉe's estimate,[1242] at least 600 fall annually from space upon the earth, ever formed part of the bodies known to us as comets. Some follow Tschermak in attributing to aerolites a totally different origin from that of periodical shooting-stars. That no clear line of demarcation can be drawn is no valid reason for asserting that no real distinction exists; and it is certainly remarkable that a meteoric fusillade may be kept up for hours without a single solid projectile reaching its destination. It would seem as if the celestial army had been supplied with blank cartridges. Yet, since a few detonating meteors have been found to proceed from ascertained radiants of shooting-stars, it is difficult to suppose that any generic difference separates them.

Their assimilation is further urged—though not with any demonstrative force—by two instances, the only two on record, of the tangible descent of an aerolite during the progress of a star-shower. On April 4, 1095, the Saxon Chronicle informs us that stars fell "so thickly that no man could count them," and adds that one of them having struck the ground in France, a bystander "cast water upon it, which was raised in steam with a great noise of boiling."[1243] And again, on November 27, 1885, while the skirts of the Andromede-tempest were trailing over Mexico, "a ball of fire" was precipitated from the sky at Mazapil, within view of a ranchman.[1244] Scientific examination proved it to be a "siderite," or mass of "nickel-iron"; its weight exceeded eight pounds, and it contained many nodules of graphite. We are not, however, authorised by the circumstances of its arrival to regard the Mazapil fragment of cosmic metal as a specimen torn from Biela's comet. In this, as in the preceding case, the coincidence of the fall with the shower may have been purely casual, since no hint is given of any sort of agreement between the tracks followed by the sample provided for curious study, and the swarming meteors consumed in the upper air.

Professor Newton's inquiries into the tracks pursued by meteorites previous to their collisions with the earth tend to distinguish them, at least specifically, from shooting-stars. He found that nearly all had been travelling with a direct movement in orbits the perihelia of which lay in the outer half of the space separating the earth from the sun.[1245] Shooting-stars, on the contrary, are entirely exempt from such limitations. The Yale Professor concluded "that the larger meteorites moving in our solar system are allied much more closely with the group of comets of short period than with the comets whose orbits are nearly parabolic." They would thus seem to be more at home than might have been expected amid the planetary family. Father Carbonelle has, moreover, shown[1246] that meteorites, if explosion-products of the earth or moon, should, with rare exceptions, follow just the kind of paths assigned to them, from data of observation, by Professor Newton. Yet it is altogether improbable that projectiles from terrestrial volcanoes should, at any geological epoch, have received impulses powerful enough to enable them, not only to surmount the earth's gravity, but to penetrate its atmosphere.

A striking—indeed, an almost startling—peculiarity, on the other hand, divides from their congeners a class of meteors identified by Mr. Denning during ten years' patient watching of such phenomena at Bristol.[1247] These are described as "meteors with stationary radiants," since for months together they seem to come from the same fixed points in the sky. Now this implies quite a portentous velocity. The direction of meteor-radiants is affected by a kind of aberration, analogous to the aberration of light. It results from a composition of terrestrial with meteoric motion. Hence, unless that of the earth in its orbit be by comparison insignificant, the visual line of encounter must shift, if not perceptibly from day to day, at any rate conspicuously from month to month. The fixity, then, of many systems observed by Mr. Denning seems to demand the admission that their members travel so fast as to throw the earth's movement completely out of the account. The required velocity would be, by Mr. Ranyard's calculation, at least 880 miles a second.[1248] But the aspect of the meteors justifies no such extravagant assumption. Their seeming swiftness is very various, and—what is highly significant—it is notably less when they pursue than when they meet the earth. Yet the "incredible and unaccountable"[1249] fact of the existence of these "long radiants," although doubted by Tisserand[1250] because of its theoretical refractoriness, must apparently be admitted. The first plausible explanation of them was offered by Professor Turner in 1899.[1251] They represent, in his view, the cumulative effects of the earth's attraction. The validity of his reasoning is, however, denied by M. BrÉdikhine,[1252] who prefers to regard them as a congeries of separate streams. The enigma they present has evidently not yet received its definitive solution.

The Perseids afford, on the contrary, a remarkable instance of a "shifting radiant." Mr. Denning's observations of these yellowish, leisurely meteors extend over nearly six weeks, from July 8 to August 16; the point of radiation meantime progressing no less than 57° in right ascension. Doubts as to their common origin were hence freely expressed, especially by Mr. Monck of Dublin.[1253] But the late Dr. Kleiber[1254] showed, by strict geometrical reasoning, that the forty-nine radiants successively determined for the shower were all, in fact, comprised within one narrowly limited region of space. In other words, the application of the proper correction for the terrestrial movement, and the effects of attraction by which each individual shooting-star is compelled to describe a hyperbola round the earth's centre, reduces the extended line of radiants to a compact group, with the cometary radiant for its central point; the cometary radiant being the spot in the sky met by a tangent to the orbit of the Perseid comet of 1862 at its intersection with the orbit of the earth. The reality of the connection between the comet and the meteors could scarcely be more clearly proved; while the vast dimensions of the stream into which the latter are found to be diffused cannot but excite astonishment not unmixed with perplexity.

The first successful application of the spectroscope to comets was by Donati in 1864.[1255] A comet discovered by Tempel, July 4, brightened until it appeared like a star somewhat below the second magnitude, with a feeble tail 30° in length. It was remarkable as having, on August 7, almost totally eclipsed a small star—a very rare occurrence.[1256] On August 5 Donati admitted its light through his train of prisms, and found it, thus analysed, to consist of three bright bands—yellow, green, and blue—separated by wider dark intervals. This implied a good deal. Comets had previously been considered, as we have seen, to shine mainly, if not wholly, by reflected sunlight. They were now perceived to be self-luminous, and to be formed, to a large extent, of glowing gas. The next step was to determine what kind of gas it was that was thus glowing in them; and this was taken by Sir William Huggins in 1868.[1257]

A comet of subordinate brilliancy, known as comet 1868 ii., or sometimes as Winnecke's, was the subject of his experiment. On comparing its spectrum with that of an olefiant-gas "vacuum tube" rendered luminous by electricity, he found the agreement exact. It has since been abundantly confirmed. All the eighteen comets tested by light analysis, between 1868 and 1880, showed the typical hydro-carbon spectrum[1258] common to the whole group of those compounds, but probably due immediately to the presence of acetylene. Some minor deviations from the laboratory pattern, in the shifting of the maxima of light from the edge towards the middle of the yellow and blue bands, have been experimentally reproduced by Vogel and Hasselberg in tubes containing a mixture of carbonic oxide with olefiant gas.[1259] Their illumination by disruptive electric discharges was, however, a condition sine qu non for the exhibition of the cometary type of spectrum. When a continuous current wasPLATE II.

employed, the carbonic oxide bands asserted themselves to the exclusion of the hydro-carbons. The distinction has great significance as regards the nature of comets. Of particular interest in this connection is the circumstance that carbonic oxide is one of the gases evolved by meteoric stones and irons under stress of heat.[1260] For it must apparently have formed part of an aeriform mass in which they were immersed at an earlier stage of their history.

In a few exceptional comets the usual carbon-bands have been missed. Two such were observed by Sir William Huggins in 1866 and 1867 respectively.[1261] In each a green ray, approximating in position to the fundamental nebular line, crossed an otherwise unbroken spectrum. And Holmes's comet of 1892 displayed only a faint prismatic band devoid of any characteristic feature.[1262] Now these three might well be set down as partially effete bodies; but a brilliant comet, visible in southern latitudes in April and May, 1901, so far resembled them in the quality of its light as to give a spectrum mainly, if not purely, continuous. This, accordingly, is no symptom of decay.

The earliest comet of first-class lustre to present itself for spectroscopic examination was that discovered by Coggia at Marseilles, April 17, 1874. Invisible to the naked eye till June, it blazed out in July a splendid ornament of our northern skies, with a just perceptibly curved tail, reaching more than half way from the horizon to the zenith, and a nucleus surpassing in brilliancy the brightest stars in the Swan. BrÉdikhine, Vogel, and Huggins[1263] were unanimous in pronouncing its spectrum to be that of marsh or olefiant gas. Father Secchi, in the clear sky of Rome, was able to push the identification even closer than had heretofore been done. The complete hydro-carbon spectrum consists of five zones of variously coloured light. Three of these only—the three central ones—had till then been obtained from comets; owing, it was supposed, to their temperature not being high enough to develop the others. The light of Coggia's comet, however, was found to contain all five, traces of the violet band emerging June 4, of the red, July 2.[1264] Presumably, all five would show universally in cometary spectra, were the dispersed rays strong enough to enable them to be seen.

The gaseous surroundings of comets are, then, largely made up of a compound of hydrogen with carbon. Other materials are also present; but the hydro-carbon element is probably unfailing and predominant. Its luminosity is, there is little doubt, an effect of electrical excitement. ZÖllner showed in 1872[1265] that, owing to evaporation and other changes produced by rapid approach to the sun, electrical processes of considerable intensity must take place in comets; and that their original light is immediately connected with these, and depends upon solar radiation, rather through its direct or indirect electrifying effects, than through its more obvious thermal power, may be considered a truth permanently acquired to science.[1266] They are not, it thus seems, bodies incandescent through heat, but glowing by electricity; and this is compatible, under certain circumstances, with a relatively low temperature.

The gaseous spectrum of comets is accompanied, in varying degrees, by a continuous spectrum. This is usually derived most strongly from the nucleus, but extends, more or less, to the nebulous appendages. In part, it is certainly due to reflected sunlight; in part, most likely, to the ignition of minute solid particles.