It is not at all easy to explain how this phantom satellite came to be seen. Father Hell, of Vienna—the same astronomer whom Sir G. Airy suspects of falling asleep during the progress of the transit of Venus in 1769—made some experiments showing how a false image of the planet might be seen beside the true one, the false image being smaller and fainter, like the moons seen by Schort (as Hell called Short), Cassini, and the rest. And more recently Sir David Brewster stated that Wargentin 'had in his possession a good achromatic telescope, which always showed Venus with such a satellite.' But Hell admitted that the falsehood of the unreal Venus was easily detected, and Brewster adds to his account of Wargentin's phantom moon, that 'the deception was discovered by turning the telescope about its axis.' As Admiral Smyth well remarks, to endeavour to explain away in this manner the observations made by Cassini and Short 'must be a mere pleasantry, for it is impossible such accurate observers could have been deceived by so gross a neglect.' Smyth, by the way, was a believer in the moon of Venus. 'The contested satellite is perhaps extremely minute,' he says, 'while some parts of its body may be less capable of reflecting light than others; and when the splendour of its primary and our inconvenient station for watching it are considered, it must be conceded that, however slight the hope may be, search ought not to be relinquished.'

Setting aside Scheuten's asserted recognition of a dark body near Venus during the transit of 1761, Venus has always appeared without any attendant when in transit. As no one else claimed to have seen what Scheuten saw in 1761, though the transit was observed by hundreds, of whom many used far finer telescopes than he, we must consider that he allowed his imagination to deceive him. During the transit of 1769, and again on December 8-9, 1874, Venus certainly had no companion during her transit.

What, then, was it that Cassini, Short, Montaigne, and the rest supposed they saw? The idea has been thrown out by Mr. Webb that mirage caused the illusion. But he appears to have overlooked the fact that though an image of Venus formed by mirage would be fainter than the planet, it would not be smaller. It might, according to the circumstances, be above Venus or below, or even somewhat towards either side, and it might be either a direct or an inverted image, but it could not possibly be a diminished image.

Single observations like Cassini's or Short's might be explained as subjective phenomena, but this explanation will not avail in the case of the Copenhagen observations.

I reject, as every student of astronomy will reject, the idea of wilful deception. Occasionally an observer may pretend to see what he has not seen, though I believe this very seldom happens. But even if Cassini and the rest had been notoriously untrustworthy persons instead of being some of them distinguished for the care and accuracy with which their observations were made and recorded, these occasional views of a phantom satellite are by no means such observations as they would have invented. No distinction was to be gained by observations which could not be confirmed by astronomers possessing more powerful telescopes. Cassini, for example, knew well that nothing but his well-earned reputation could have saved him from suspicion or ridicule when he announced that he had seen Venus attended by a satellite.

It seems to me probable that the false satellite was an optical illusion brought about in a different way from those referred to by Hell and Brewster, though among the various circumstances which in an imperfect instrument might cause such a result I do not undertake to make a selection. It is certain that Venus's satellite has vanished with the improvement of telescopes, while it is equally certain that even with the best modern instruments illusions occasionally appear which deceive even the scientific elect. Three years have passed since I heard the eminent observer Otto Struve, of Pulkowa, give an elaborate account of a companion to the star Procyon, describing the apparent brightness, distance, and motions of this companion body, for the edification of the Astronomer-Royal and many other observers. I had visited but a few months before the Observatory at Washington, where, with a much more powerful telescope, that companion to Procyon had been systematically but fruitlessly sought for, and I entertained a very strong opinion, notwithstanding the circumstantial nature of Struve's account and his confidence (shared in unquestioningly by the observers present), that he had been in some way deceived. But I could not then see, nor has any one yet explained, how this could be. The fact, however, that he had been deceived is now undoubted. Subsequent research has shown that the Pulkowa telescope, though a very fine instrument, possesses the undesirable quality of making a companion orb for all first-class stars in the position where O. Struve and his assistant Lindenau saw the supposed companion of Procyon.

I may as well point out, however, that theories so wild have recently been broached respecting Venus, that far more interesting explanations of the enigma than this optical one may be looked for presently. It has been gravely suggested by Mr. Jos. Brett, the artist, that Venus has a surface of metallic brilliancy, with a vitreous atmosphere,—which can only be understood to signify a glass case. This stupendous theory has had its origin in an observation of considerable interest which astronomers (it is perhaps hardly necessary to say) explain somewhat differently. When Venus has made her entry in part upon the sun's face at the beginning of transit, there is seen all round the portion of her disc which still remains outside the sun an arc of light so brilliant that it records its photographic trace during the instantaneous exposure required in solar photography. It is mathematically demonstrable that this arc of light is precisely what should be seen if Venus has an atmosphere like our earth's. But mathematical demonstration is not sufficient (or perhaps we may say it is too much) for some minds. Therefore, to simplify matters, Venus has been provided with a mirror surface and a glass case. (See preceding essay, on Astronomical Paradoxes, for further details.)

The enigma next to be considered is of a more doubtful character than the myth relating to the satellite of Venus. Astronomers are pretty well agreed that Venus has no moon, but many, including some deservedly eminent, retain full belief in the story of the planet Vulcan.

More than seventeen years ago the astronomical world was startled by the announcement that a new planet had been discovered, under circumstances unlike any which had heretofore attended the discovery of fresh members of the solar system. At that time astronomers had already become accustomed to the discovery, year after year, of several asteroids, which are in reality planets, though small ones. In fact, no less than fifty-six of these bodies were then known, whereof fifty-one had been discovered during the years 1847-1858 inclusive, not one of these years having passed without the detection of an asteroid. But all these planets belonged to one family, and as there was every reason to believe that thousands more travel in the same region of the solar system, the detection of a few more among the number had no longer any special interest for astronomers. The discovery of the first known member of the family had indeed been full of interest, and had worthily inaugurated the present century, on the first day of which it was made. For it had been effected in pursuance of a set scheme, and astronomers had almost given up all hopes of success in that scheme when Piazzi announced his detection of little Ceres. Again the discovery of the next few members of the family had been interesting as revealing the existence of a new order of bodies in the solar system. No one had suspected the possibility that besides the large bodies which travel round the sun, either singly or attended by subordinate families of moons, there might be a ring of many planets. This was what the discovery of Ceres, Pallas, Juno, and Vesta seemed to suggest, unless—still stranger thought—these were but fragments of a mighty planet which had been shattered in long-past ages by some tremendous explosion. Since then, however, this startling theory has been (itself) exploded. Year after year new members of the ring of multitudinous planets are discovered, and that, not as was recently predicted, in numbers gradually decreasing, but so rapidly that more have been discovered during the last ten years than during the preceding twenty.

The discovery of the giant planet Uranus, an orb exceeding our earth twelve and a half times in mass and seventy-four times in volume, was a matter of much greater importance, so far as the dignity of the planetary system was concerned, for it is known that the whole ring of asteroids together does not equal one-tenth part of the earth in mass, while Uranus exceeds many times in volume the entire family of terrestrial planets—Mercury, Venus, the Earth, and Mars. The detection of Uranus, unlike that of Ceres, was effected by accident. Sir W. Herschel was looking for double stars of a particular kind in the constellation Gemini when by good fortune the stranger was observed.

The interest with which astronomers received the announcement of the discovery of Uranus, though great, was not to be compared with that with which they deservedly welcomed the discovery of Neptune, a larger and more massive planet, revolving at a distance one-half greater even than the mighty space which separates Uranus from the sun, a space so great that by comparison with it the range of 184,000,000 of miles, which forms the diameter of our earth's orbit, seems quite insignificant. It was not, however, the vastness of Neptune's mass or volume, or the awful remoteness of the path along which he pursues his gloomy course, which attracted the interest of astronomers, but the strangeness of the circumstances under which the planet had been detected. His influence had been felt for many years before astronomers thought of looking for him, and even when the idea had occurred to one or two, it was considered, and that, too, by an astronomer as deservedly eminent as Sir G. Airy, too chimerical to be reasonably entertained. All the world now knows how Leverrier, the greatest living master of physical astronomy, and Adams, then scarce known outside Cambridge, both conceived the idea of finding the planet, not by the simple method of looking for it with a telescope, but by the mathematical analysis of the planet's disturbing influence upon known members of the solar system. All know, too, that these mathematicians succeeded in their calculations, and that the planet was found in the very region and close to the very point indicated first by Adams, and later, but independently, and (fortunately for him more publicly) by Leverrier.

None of these instances of the discovery of members of the solar system resembled in method or details the discovery announced early in the year 1859. It was not amid the star-depths and in the darkness of night that the new planet was looked for, but in broad day, and on the face of the sun himself. It was not on the outskirts of the solar system that the planet was supposed to be travelling, but within the orbit of Mercury, hitherto regarded as of all planets the nearest to the sun. It was not hoped that any calculation of the perturbations of other planets would show the place of the stranger, though certain changes in the orbit of Mercury seemed clearly enough to indicate the stranger's existence.

Early in 1860 Leverrier had announced that the position of Mercury's path was not precisely in agreement with calculations based on the adopted estimates of the masses of those planets which chiefly disturb the motions of Mercury. The part of the path where Mercury is nearest to the sun, and where, therefore, he travels fastest, had slightly shifted from its calculated place. This part of the path was expected to move, but it had moved more than was expected; and of course Mercury having his region of swiftest motion somewhat differently placed than was anticipated, himself moved somewhat differently.

Leverrier found that to explain this feature of Mercury's motion either the mass of Venus must be regarded as one-tenth greater than had been supposed, or some unknown cause must be regarded as affecting the motion of Mercury. A planet as large as Mercury, about midway between Mercury and the sun, would account for the observed disturbance; but Leverrier rejected the belief that such a planet exists, simply because he could not 'believe that it would be invisible during total eclipses of the sun.' 'All difficulties disappear,' he added, 'if we admit, in place of a single planet, small bodies circulating between Mercury and the sun.' Considering their existence as not at all improbable, he advised astronomers to watch for them.

It was on January 2, 1860, that Leverrier thus wrote. On December 22, 1859, a letter had been addressed by a M. Lescarbault of Orgeres to Leverrier, through M. Vallee, hon. inspector-general of roads and bridges, announcing that on March 26, 1859, about four in the afternoon, Lescarbault had seen a round black spot on the face of the sun, and had watched it as it passed across like a planet in transit—not with the slow motion of an ordinary sun-spot. The actual time during which the round spot was visible was one hour, seventeen minutes, nine seconds, the rate of motion being such that, had the spot crossed the middle of the sun's disc, at the same rate, the transit would have lasted more than four hours. The spot thus merely skirted the sun's disc, being at no time more than about one forty-sixth part of the sun's apparent diameter from the edge of the sun. Lescarbault expressed his conviction that on a future day, a black spot, perfectly round and very small, will be seen passing over the sun, and 'this point will very probably be the planet whose path I observed on March 26, 1859.' 'I am persuaded,' he added, 'that this body is the planet, or one of the planets, whose existence in the vicinity of the sun M. Leverrier had made known a few months ago' (referring to the preliminary announcement of results which Leverrier published afterwards more definitely).

Leverrier, when the news of Lescarbault's observation first reached him, was surprised that the observation should not have been announced earlier. He did not consider the delay sufficiently justified by Lescarbault's statement that he wished to see the spot again. He therefore set out for Orgeres, accompanied by M. Vallee. 'The predominant feeling in Leverrier's mind,' says Abbe Moigno, 'was the wish to unmask an attempt to impose upon him, as the person more likely than any other astronomer to listen to the allegation that his prophecy had been fulfilled.'

'One should have seen M. Lescarbault,' says Moigno, 'so small, so simple, so modest, and so timid, in order to understand the emotion with which he was seized, when Leverrier, from his great height, and with that blunt intonation which he can command, thus addressed him: "It is then you, sir, who pretend to have observed the intra-mercurial planet, and who have committed the grave offence of keeping your observation secret for nine months. I warn you that I have come here with the intention of doing justice to your pretensions, and of demonstrating either that you have been dishonest or deceived. Tell me, then, unequivocally, what you have seen."' This singular address did not bring the interview, as one might have expected, to an abrupt end. The lamb, as the Abbe calls the doctor, trembling, stammered out an account of what he had seen. He explained how he had timed the passage of the black spot. 'Where is your chronometer?' asked Leverrier. 'It is this watch, the faithful companion of my professional journeys.' 'What! with that old watch, showing only minutes, dare you talk of estimating seconds. My suspicions are already too well confirmed.' 'Pardon me, I have a pendulum which beats seconds.' 'Show it me.' The doctor brings down a silk thread to which an ivory ball is attached. Fixing the upper end to a nail, he draws the ball a little from the vertical, counts the number of oscillations, and shows that his pendulum beats seconds; he explains also how his profession, requiring him to feel pulses and count pulsations, he has no difficulty in mentally keeping record of successive seconds.

Having been shown the telescope with which the observation was made, the record of the observation (on a piece of paper covered with grease and laudanum, and doing service as a marker in the 'Connaissance des Temps,' or French Nautical Almanac), Leverrier presently inquired if Lescarbault had attempted to deduce the planet's distance from the sun from the period of its transit. The doctor admitted that he had attempted this, but, being no mathematician, had failed to achieve success with the problem. He showed the rough draughts of his futile attempts at calculation on a board in his workshop, 'for,' said he naively, 'I am a joiner as well as an astronomer.'

The interview satisfied Leverrier that a new planet, travelling within the orbit of Mercury, had really been discovered. 'With a grace and dignity full of kindness,' says a contemporary narrative of these events,[55] 'he congratulated Lescarbault on the important discovery which he had made.' Anxious to obtain some mark of respect for the discoverer of Vulcan, Leverrier made inquiry concerning his private character, and learned from the village cure, the juge de paix, and other functionaries, that he was a skilful physician and a worthy man. With such high recommendations, M. Leverrier requested from M. Rouland, the Minister of Public Instruction, the decoration of the Legion of Honour for M. Lescarbault. The Minister, in a brief but interesting statement of his claim, communicated this request to the Emperor, who, by a decree dated January 25, conferred upon the village astronomer the honours so justly due to him. His professional brethren in Paris were equally solicitous to testify their regard; and MM. Felix Roubaud, Legrande, and Caffe, as delegates of the scientific press, proposed to the medical body, and to the scientific world in Paris, to invite Lescarbault to a banquet in the Hotel du Louvre on January 18.

The announcement of the supposed discovery caused astronomers to re-examine records of former observations of black spots moving across the sun. Several such records existed, but they had gradually come to be regarded as of no real importance. Wolff of Zurich published a list of no fewer than twenty such observations made since 1762. Carrington added many other cases. Comparing together three of these observations, Wolff found that they would be satisfied by a planet having a period of revolution of 19 days, agreeing fairly with the period of rather more than 19-1/3 days inferred by Leverrier for Lescarbault's planet. But the entire set of observations of black spots require that there should be at least three new planets travelling between Mercury and the sun. Many observers also set themselves the task of searching for Vulcan, as the supposed new planet was called. They have continued fruitlessly to observe the sun for this purpose until the present time.

While the excitement over Lescarbault's discovery was at its height, another observer impugned not only the discovery but the honesty of the discoverer.

M. Liais, a French astronomer of considerable skill, formerly of the Paris Observatory, but at the time of Lescarbault's achievement in the service of the Brazilian Government, published a paper, 'Sur la Nouvelle Planete annoncee par M. Lescarbault,' in which he endeavoured to establish the four following points:—

First, the observation of Lescarbault was never made.

Secondly, Leverrier was mistaken in considering that a planet such as Vulcan might have escaped detection when off the sun's face.

Thirdly, that Vulcan would certainly have been seen during total solar eclipses, if the planet had a real objective existence.

Fourthly, M. Leverrier's reasons for believing that the planet exists are based on the supposition that astronomical observations are more precise than they really are.

Probably, Liais's objections would have had more weight with Leverrier had the fourth point been omitted. It was rash in a former subordinate to impugn the verdict of the chief of the Paris Observatory on a matter belonging to that special department of astronomy which an observatory chief might be expected to understand thoroughly. It is thought daring in the extreme for one outside the circles of official astronomy (as Newton in Flamstead's time, Sir W. Herschel in Maskelyne's, and Sir J. Herschel in the present century), to advance or maintain an opinion adverse to that of some official chief, but for a subordinate (even though no longer so), to be guilty of such rash procedure 'is most tolerable and not to be endured,' as a typical official has said. Accordingly, very little attention was paid by Leverrier to Liais's objections.

Yet, in some respects, what M. Liais had to say was very much to the point.

At the very time when Lescarbault was watching the black spot on the sun's face, Liais was examining the sun with a telescope of much greater magnifying power, and saw no such spot. His attention was specially directed to the edge of the sun (where Lescarbault saw the spot) because he was engaged in determining the decrease of the sun's brightness near the edge. Moreover, he was examining the very part of the sun's edge where Lescarbault saw the planet enter, at a time when it must have been twelve minutes in time upon the face of the sun, and well within the margin of the solar disc. The negative evidence here is strong; though it must always be remembered that negative evidence requires to be overwhelmingly strong before it can be admitted as effective against positive evidence. It seems at a first view utterly impossible that Liais, examining with a more powerful telescope the region where Lescarbault saw the spot, could have failed to see it had it been there; but experience shows that it is not impossible for an observer engaged in examining phenomena of one class to overlook a phenomenon of another class, even when glaringly obvious. All we can say is that Liais was not likely to have overlooked Lescarbault's planet had it been there; and we must combine this probability against Vulcan's existence with arguments derived from other considerations. There is also the possibility of an error in time. As the writer in the 'North British Review' remarks, 'twelve minutes is so short a time that it is just possible that the planet may not have entered upon the sun during the time that Liais observed it.'

The second and third arguments are stronger. In fact, I do not see how they can be resisted.

It is, in the first place, clear from Lescarbault's account that Vulcan must have a considerable diameter—certainly if Vulcan's diameter in miles were only half the diameter of Mercury, it would have been all but impossible for Lescarbault with his small telescope to see Vulcan at all, whereas he saw the black spot very distinctly. Say Vulcan has half the diameter of Mercury, and let us compare the brightness of these two planets when at their greatest apparent distances from the sun, that is, when each looks like a half-moon. The distance of Mercury exceeds the estimated distance of Vulcan from the sun as 27 exceeds 10, so that Vulcan is more strongly illuminated in the proportion of 27 times 27 to 10 times 10, or 729 to 100—say at least 7 to 1. But having a diameter but half as large the disc of Vulcan could be but about a fourth of Mercury's at the same distance from us (and they would be at about the same distance from us when seen as half-moons). Hence Vulcan would be brighter than Mercury in the proportion of 7 to 4. Of course being so near the sun he would not be so easily seen; and we could never expect to see him at all, perhaps, with the naked eye—though even this is not certain. But Mercury, when at the same apparent distance from the sun, and giving less light than at his greatest seeming distance, is quite easily seen in the telescope. Much more easily, then, should Vulcan be seen, if a telescope were rightly directed at such a time, or when Vulcan was anywhere near his greatest seeming distance from the sun. Now it is true astronomers do not know precisely when or where to look for him. But he passes from his greatest distance on one side of the sun to his greatest distance on the other in less than ten days, according to the computed period, and certainly (that is, if the planet exists) in a very short time. The astronomer has then only to examine day after day a region of small extent on either side of the sun, for ten or twelve days in succession (an hour's observation each day would suffice), to be sure of seeing Vulcan. Yet many astronomers have made such search many times over, without seeing any trace of the planet. During total solar eclipses, again, the planet has been repeatedly looked for unsuccessfully—though it should at such a time be a very conspicuous object, when favourably placed, and could scarcely fail of being very distinctly seen wherever placed.

The fourth argument of Lescarbault's is not so effective, and in fact he gets beyond his depth in dealing with it. But it is to be noticed that a considerable portion of the discrepancy between Mercury's observed and calculated motions has long since been accounted for by the changed estimate of the earth's mass as compared with the sun's, resulting from the new determination of the sun's distance. However, the arguments depending on this consideration would not be suited to these pages.

There was one feature in Liais's paper which was a little unfortunate. He questioned Lescarbault's honesty. He said 'Lescarbault contradicts himself in having first asserted that he saw the planet enter upon the sun's disc, and having afterwards admitted to Leverrier that it had been on the disc some seconds before he saw it, and that he had merely inferred the time of its entry from the rate of its motion afterwards. If this one assertion be fabricated, the whole may be so.' 'He considers these arguments to be strengthened,' says the 'North British Review,' 'by the assertion which, as we have seen, perplexed Leverrier himself, that if M. Lescarbault had actually seen a planet on the sun, he could not have kept it secret for nine months.'

This charge of dishonesty, unfortunate in itself, had the unfortunate effect of preventing Lescarbault or the Abbe Moigno from replying. The latter simply remarked that the accusation was of such a nature as to dispense him from any obligation to refute it. This was an error of judgment, I cannot but think, if an effective reply was really available.

The Remarks with which the North British Reviewer closes his account may be repeated now, so far as they relate to the force of the negative evidence, with tenfold effect. 'Since the first notice of the discovery in the beginning of January 1860 the sun has been anxiously observed by astronomers; and the limited area around him in which the planet must be, if he is not upon the sun, has doubtless been explored with equal care by telescopes of high power, and processes by which the sun's direct light has been excluded from the tube of the telescope as well as the eye of the observer, and yet no planet has been found. This fact would entitle us to conclude that no such planet exists if its existence had been merely conjectured, or if it had been deduced from any of the laws of planetary distance, or even if Leverrier or Adams had announced it as the probable result of planetary perturbations. If the finest telescopes cannot rediscover a planet which with the small power used by Lescarbault has a visible disc, within so limited an area of which the sun is the centre, or rather within a narrow belt of that circle, we should unhesitatingly declare that no such planet exists. But the question assumes a very different aspect when it involves moral considerations. If,' proceeds the Reviewer, writing in August 1860, 'after the severe scrutiny which the sun and its vicinity will undergo before and after and during his total eclipse in July, no planet shall be seen; and if no round black spot distinctly separable from the usual solar spots shall be seen on the solar spots' (sic, presumably solar disc was intended), 'we will not dare to say that it does not exist. We cannot doubt the honesty of M. Lescarbault, and we can hardly believe that he was mistaken. No solar spot, no floating scoria, could maintain in its passage over the sun a circular and uniform shape, and we are confident that no other hypothesis but that of an intra-mercurial planet can explain the phenomena seen and measured by M. Lescarbault, a man of high character, possessing excellent instruments, and in every way competent to use them well, and to describe clearly and correctly the results of his observations. Time, however, tries facts as well as speculations. The phenomena observed by the French astronomer may never be again seen, and the disturbance of Mercury which rendered it probable may be otherwise explained. Should this be the case, we must refer the round spot on the sun to some of those illusions of the eye or of the brain which have sometimes disturbed the tranquillity of science.'

The evidence which has accumulated against Vulcan in the interval since this was written is not negative only, but partly positive, as the following instance, which I take from my own narrative at the time in a weekly journal, serves to show:—After more than sixteen years of fruitless watching, astronomers learned last August (1876) that in the month of April Vulcan had been seen on the sun's disc in China. On April 4, it appeared, Herr Weber, an observer of considerable skill, stationed at Pecheli, had seen a small round spot on the sun, looking very much as a small planet might be expected to look. A few hours later he turned his telescope upon the sun, and lo! the spot had vanished, precisely as though the planet had passed away after the manner of planets in transit. He forwarded the news of his observation to Europe. The astronomer Wolff, well known for his sun-spot studies, carefully calculated the interval which had passed since Lescarbault saw Vulcan on March 26, 1859, and to his intense satisfaction was enabled to announce that this interval contained the calculated period of the planet an exact number of times. Leverrier at Paris received the announcement still more joyfully; while the Abbe Moigno, who gave Vulcan its name, and has always staunchly believed in the planet's existence, congratulated Lescarbault warmly upon this new view of the shamefaced Vulcan. Not one of those who already believed in the planet had the least doubt as to the reality of Weber's observations, and of these only Lescarbault himself received the news without pleasure. He, it seems, has never forgiven the Germans for destroying his observatory and library during the invasion of France in 1870, and apparently would prefer that his planet should never be seen again rather than that a German astronomer should have seen it. But the joy of the rest and Lescarbault's sorrow were alike premature. It was found that the spot seen by Weber had not only been observed at the Madrid observatory, where careful watch is kept upon the sun, but had been photographed at Greenwich; and when the description of its appearance, as seen in a powerful telescope at one station, and its picture as photographed by a fine telescope at the other, came to be examined, it was proved unmistakably that the spot was an ordinary sun-spot (not even quite round), which had after a few hours disappeared, as even larger sun-spots have been known to do in even a shorter time.

It is clear that had not Weber's spot been fortunately seen at Madrid and photographed at Greenwich, his observation would have been added to the list of recorded apparitions of Vulcan in transit, for it fitted in perfectly with the theory of Vulcan's real existence. I think, indeed, for my own part, that the good fortune was Weber's. Had it so chanced that thick weather in Madrid and at Greenwich had destroyed the evidence actually obtained to show that what Weber described he really saw, although it was not what he thought, some of the more suspicious would have questioned whether, in the euphonious language of the North British Reviewer, 'the round spot on the sun' was not due 'to one of those illusions of the eye or of the brain which have sometimes disturbed the tranquillity of science.' Of course no one acquainted with M. Weber's antecedents would imagine for a moment that he had invented the observation, even though the objective reality of his spot had not been established. But if a person who is entirely unknown, states that he has seen Vulcan, there is antecedently some degree of probability in favour of the belief that the observation is as much a myth as the planet itself. Some observations of Vulcan have certainly been invented. I have received several letters purporting to describe observations of bodies in transit over the sun's face, either the rate of transit, the size of the body, or the path along which it was said to move, being utterly inconsistent with the theory that it was an intra-mercurial planet, while yet (herein is the suspicious circumstance of such narratives) the epoch of transit accorded in the most remarkable manner with the period assigned to Vulcan. A paradoxist in America (of Louisville, Kentucky) who had invented a theory of the weather, in which the planets, by their influence on the sun, were supposed to produce all weather-changes, the nearer planets being the most effective, found his theory wanted Vulcan very much. Accordingly, he saw Vulcan crossing the sun's face in September, which, being half a year from March, is a month wherein, according to Lescarbault's observation, Vulcan may be seen in transit, and by a strange coincidence the interval between our paradoxist's observation and Lescarbault's exactly contained a certain number of times the period calculated by Leverrier for Vulcan. This was a noble achievement on the part of our paradoxist. At one stroke it established his theory of the weather, and promised to ensure him text-book immortality as one of the observers of Vulcan. But, unfortunately, a student of science residing in St. Louis, after leaving the Louisville paradoxist full time to parade his discovery, heartlessly pointed out that an exact number of revolutions of Vulcan after Lescarbault's March observation, must of necessity have brought the planet on that side of the sun on which the earth lies in March, so that to see Vulcan so placed on the sun's face in September was to see Vulcan through the sun, a very remarkable achievement indeed. The paradoxist was abashed, the reader perhaps imagines. Not in the least. The planet's period must have been wrongly calculated by Leverrier—that was all: the real period was less than half as long as Leverrier had supposed; and instead of having gone a certain number of times round since Lescarbault had seen it, Vulcan had gone twice as many times round and half once round again. The circumstance that if Vulcan's period had been thus short, the time of crossing the sun's face would have been much less than, according to Lescarbault's account, it actually was, had not occurred to the Louisville weather-prophet.[56]

Leverrier's faith in Vulcan, however, has remained unshaken. He has used all the observations of spots which, like Weber's, have been seen only for a short time. At least he has used all which have not, like Weber's, been proved to be only transient sun-spots. Selecting those which fit in well with Lescarbault's observation, he has pointed out how remarkable it is that they show this accord. The possibility that some of them might be explicable as Weber's proved to be, and that some even may have been explicable as completely, but less satisfactorily, in another way, seems to have been thought scarcely worth considering. Using the imperfect materials available, but with exquisite skill—as a Phidias might model an exquisite figure of materials that would presently crumble into dust—Leverrier came to the conclusion that Vulcan would cross the sun's disc on or about March 22, 1876. 'He, therefore,' said Sir G. Airy, addressing the Astronomical Society, 'circulated a despatch among his friends, asking them carefully to observe the sun on March 22.' Sir G. Airy, humouring his honoured friend, sent telegrams to India, Australia, and New Zealand, requesting that observations might be made every two hours or oftener. Leverrier himself wrote to Santiago de Chili and other places, so that, including American and European observations, the sun could be watched all through the twenty-four hours on March 21, 22, and 23. 'Without saying positively that he believed or disbelieved in the existence of the planet,' proceeds the report, 'Sir G. Airy thought, since M. Leverrier was so confident, that the opportunity ought not to be neglected by anybody who professed to take an interest in the progress of planetary astronomy.'

It is perhaps unnecessary to add that observations were made as requested. Many photographs of the sun also were taken during the hours when Vulcan, if he exists at all, might be expected to cross the sun's face. But the 'planet of romance,' as Abbe Moigno has called Vulcan, failed to appear, and the opinion I had expressed last October ('English Mechanic and World of Science,' for October 27, 1876, p. 160), that Vulcan might perhaps better be called the 'planet of fiction' was pro tanto confirmed. Nevertheless, I would not be understood to mean by the word 'fiction' aught savouring of fraud so far as Lescarbault is concerned—I prefer the North Briton's view of Lescarbault's spot, that so to speak, it was

... the blot upon his brain, That would show itself without.

I have left small space to treat of other fancied discoveries among the orbs of heaven. Yet there are some which are not only interesting but instructive, as showing how even the most careful observers may be led astray. In this respect the mistakes into which observers of great and well deserved eminence have fallen are specially worthy of attention. With the description of three such mistakes, made by no less an astronomer than Sir W. Herschel, I shall bring this paper to a close.

When Sir W. Herschel examined the planet Uranus with his most powerful telescope he saw the planet to all appearance girt about by two rings at right angles to one another. The illusion was so complete that Herschel for several years remained in the belief that the rings were real. They were, however, mere optical illusions, due to the imperfect defining qualities of the telescope with which he observed the planet. Later he wrote that 'the observations which tend to ascertain' (indicate?) 'the existence of rings not being satisfactorily supported, it will be proper that surmises of them should either be given up, as ill-founded, or at least reserved till superior instruments can be provided.'

Sir W. Herschel was more completely misled by the false Uranian satellites. He had seen, as he supposed, no less than six of these bodies. As only two of these had been seen again, while two more were discovered by Lassell, the inference was that Uranus has eight satellites in all. These for a long time flourished in our text-books of astronomy; and many writers, confident in the care and skill of Sir W. Herschel, were unable for a long time to believe that he had been deceived. Thus Admiral Smyth, in his 'Celestial Cycle,' wrote of those who doubted the extra satellites:—'They must have but a meagre notion of Sir W. Herschel's powerful means, his skill in their application, and his method of deliberate procedure. So far from doubting there being six satellites' (this was before Lassell had discovered the other two) 'it is highly probable that there are still more.' Whewell, also, in his 'Bridgewater Treatise,' says, 'that though it no longer appears probable that Uranus has a ring like Saturn, he has at least five satellites which are visible to us, and we believe that the astronomer will hardly deny that he' (Uranus, not the astronomer), 'may possibly have thousands of smaller ones circulating about him.' But in this case Sir W. Herschel, anxiously though he endeavoured to guard against the possibility of error, was certainly mistaken. Uranus may, for anything that is known to the contrary, have many small satellites circulating about him, but he certainly has not four satellites (besides those known) which could have been seen by Sir W. Herschel with the telescope he employed. For the neighbourhood of the planet has been carefully examined with telescopes of much greater power by observers who with those telescopes have seen objects far fainter than the satellites supposed to have been seen by the elder Herschel.

The third of the Herschelian myths was the lunar volcano in eruption, which he supposed he had seen in progress in that part of the moon which was not at the time illuminated by the sun's rays. He saw a bright star-like point of light, which corresponded in position with the crater of the lunar mountain Aristarchus. He inferred that a volcano was in active eruption because the brightness of the point of light varied from time to time, and also because he did not remember to have seen it before under the same conditions. There is no doubt something very remarkable in the way in which this part of the moon's surface shines when not illumined by the sun. If it were always bright we should conclude at once that the earth-light shining upon it rendered it visible. For it must be remembered that the part of the moon which looks dark (or seems wanting to the full disc) is illuminated by our earth, shining in the sky of the moon as a disc thirteen times as large as that of the moon we see, and with the same proportion of its disc sunlit as is dark in the moon's disc. Thus when the moon is nearly new our earth is shining in the lunar skies as a nearly full moon thirteen times as large as ours. The light of this noble moon must illumine the moon's surface much more brightly than a terrestrial landscape is illumined by the full moon, and if any parts of her surface are very white they will shine out from the surface around, just as the snow-covered peak of a mountain shines out upon a moonlit night from among the darker hills and dales and rocks and forests of the landscape. But Herschel considered that the occasional brightness of the crater Aristarchus could not be thus explained. The spot had been seen before the time of Herschel's observations by Cassini and others. It has been seen since by Captain Kater, Francis Baily, and many others. Dr. Maskelyne tells us that in March 1794 it was seen by the naked eye by two persons.

Baily thus describes the appearance presented by this lunar crater on December 22, 1835: 'Directed telescope to the moon, and pointing it to the dark part in the vicinity of Aristarchus, soon saw the outline of that mountain very distinctly, formed like an irregular nebula. Nearly in the centre was a light resembling that of a star of the ninth or tenth magnitude. It appeared by glimpses, but at times was brilliant, and visible for several seconds together.'

There can be little doubt, however, that the apparent brightness of this lunar crater, or rather of its summit, is due to some peculiar quality in the surface, which may perhaps be covered by some crystalline or vitreous matter poured out in the far distant time when the crater was an active one. Prof. Shaler, who examined the crater when it was illuminated only by earthshine, with the fine 15-inch telescope of the Harvard Observatory (Cambridge U.S.), says that he has been able to recognise nearly all the craters over 15 miles in diameter in the dark part. 'There are several degrees of brightness,' he says, 'observable in the different objects which shine out by the earth-light. This fact probably explains the greater part of the perplexing statements concerning the illumination of certain craters. It certainly accounts for the volcanic activity which has so often been supposed to be manifested by Aristarchus. Under the illumination by the earth-light this is by far the brightest object on the dark part of the moon's face, and is visible much longer and with poorer glasses than any other object there.'

Here my record of astronomical myths must be brought to a close. It will be noticed that in every instance either the illusion has affected the actual observations of eminent and skilful astronomers, or has caused such astronomers to put faith for a while in illusory observations. Had I cared to include the mistakes which have been made by or have misled observers of less experience, I could have filled many sheets for each page of the present article. But it has seemed to me more instructive to show how error may affect the observations even of the most careful and deservedly eminent astronomers, how even the most cautious may be for a time misled by the mistakes of inferior observers, especially when the fact supposed to have been observed accords with preconceived opinions.



XII.

THE ORIGIN OF THE CONSTELLATION-FIGURES.

Although the strange figures which astronomers still allow to straggle over their star maps no longer have any real scientific interest, they still possess a certain charm, not only for the student of astronomy, but for many who care little or nothing about astronomy as a science. When I was giving a course of twelve lectures in Boston, America, a person of considerable culture said to me, 'I wish you would lecture about the constellations; I care little about the sun and moon and the planets, and not much more about comets; but I have always felt great interest in the Bears and Lions, the Chained and Chaired Ladies, King Cepheus and the Rescuer Perseus, Orion, Ophiuchus, Hercules, and the rest of the mythical and fanciful beings with which the old astronomers peopled the heavens. I say with Carlyle, "Why does not some one teach me the constellations, and make me at home in the starry heavens, which are always overhead, and which I don't half know to this day."' We may notice, too, that the poets by almost unanimous consent have recognised the poetical aspect of the constellations, while they have found little to say about subjects which belong especially to astronomy as a science. Milton has indeed made an Archangel reason (not unskilfully for Milton's day) about the Ptolemaic and Copernican systems, while Tennyson makes frequent reference to astronomical theories. 'There sinks the nebulous star we call the Sun, if that hypothesis of theirs be sound,' said Ida; but she said no more, save 'let us down and rest,' as though the subject were wearisome to her. Again, in the Palace of Art the soul of the poet having built herself that 'great house so royal, rich, and wide,' thither—

... when all the deep unsounded skies Shuddered with silent stars, she clomb, And as with optic glasses her keen eyes Pierced through the mystic dome, Regions of lucid matter taking forms, Brushes of fire, hazy gleams, Clusters and beds of worlds and beelike swarms Of suns, and starry streams: She saw the snowy poles of moonless Mars, That marvellous round of milky light Below Orion, and those double stars Whereof the one more bright Is circled by the other.

But the poet's soul so wearied of these astronomical researches that the beautiful lines I have quoted disappeared (more's the pity) from the second and all later editions. Such exceptions, indeed, prove the rule. Poets have been chary in referring to astronomical researches and results, full though these have been of unspeakable poetry; while from the days of Homer to those of Tennyson, the constellations which 'garland the heavens' have always been favourite subjects of poetic imagery.

It is not my present purpose, however, to discuss the poetic aspect of the constellations. I propose to inquire how these singular figures first found their way to the heavens, and, so far as facts are available for the purpose, to determine the history and antiquity of some of the more celebrated constellations.

Long before astronomy had any existence as a science men watched the stars with wonder and reverence. Those orbs, seemingly countless—which bespangle the dark robe of night—have a charm and beauty of their own apart from the significance with which the science of astronomy has invested them. The least fanciful mind is led to recognise on the celestial concave the emblems of terrestrial objects, pictured with more or less distinctness among the mysterious star-groupings. We can imagine that long before the importance of the study of the stars was recognised, men had begun to associate with certain star-groups the names of familiar objects animate or inanimate. The flocks and herds which the earliest observers of the heavens tended would suggest names for certain sets of stars, and thus the Bull, the Ram, the Kids, would appear in the heavens. Other groups would remind those early observers of the animals from whom they had to guard their flocks, or of the animals to whose vigilance they trusted for protection, and thus the Bear, the Lion, and the Dogs would find their place among the stars. The figures of men and horses, and of birds and fishes, would naturally enough be recognised, nor would either the implements of husbandry, or the weapons by which the huntsman secured his prey, remain unrepresented among the star-groupings. And lastly, the altar on which the first-fruits of harvest and vintage were presented, or the flesh of lambs and goats consumed, would be figured among the innumerable combinations which a fanciful eye can recognise among the orbs of heaven.

In thus suggesting that the first observers of the heavens were shepherds, huntsmen, and husbandmen, I am not advancing a theory on the difficult questions connected with the origin of exact astronomy. The first observations of the heavens were of necessity made by men who depended for their subsistence on a familiarity with the progress and vicissitudes of the seasons, and doubtless preceded by many ages the study of astronomy as a science. And yet the observations made by those early shepherds and hunters, unscientific though they must have been in themselves, are full of interest to the student of modern exact astronomy. The assertion may seem strange at first sight, but is nevertheless strictly true, that if we could but learn with certainty the names assigned to certain star-groups, before astronomy had any real existence, we could deduce lessons of extreme importance from the rough observations which suggested those old names. In these days, when observations of such marvellous exactness are daily and nightly made, when instruments capable of revealing the actual constitution of the stars are employed, and observers are so numerous, it may seem strange to attach any interest to the question whether half-savage races recognised in such and such a star-group the likeness of a bear, or in another group the semblance of a ship. But though we could learn more, of course, from exacter observations, yet even such rough and imperfect records would have their value. If we could be certain that in long-past ages a star-group really resembled some known object, we should have in the present resemblance of that group to the same object evidence of the general constancy of stellar lustre, or if no resemblance could be recognised we should have reason to doubt whether other suns (and therefore our own sun) may not be liable to great changes.

The subject of the constellation-figures as first known is interesting in other ways. For instance, it is full of interest to the antiquarian (and most of us are to some degree antiquarians) as relating to the most ancient of all human sciences. The same mental quality which causes us to look with interest on the buildings raised in long-past ages, or on the implements and weapons of antiquity, renders the thought impressive that the stars which we see were gazed on perhaps not less wonderingly in the very infancy of the human race. It is, again, a subject full of interest to the chronologist to inquire in what era of the world's history exact astronomy began, the moon was assigned her twenty-eight zodiacal mansions, the sun his twelve zodiacal signs. It is well known, indeed, that Newton himself did not disdain to study the questions thus suggested; and the speculations of the ingenious Dupuis found favour with the great mathematician Laplace.

Unfortunately, the evidence is not sufficiently exact to be very trustworthy. In considering, for instance, the chronological inquiries of Newton, one cannot but feel that the reliance placed by him on the statements made by different writers is not justified by the nature of those statements, which were for the most part vague in the extreme. We owe many of them to poets who, knowing little of astronomy, mixed up the phenomena of their own time with those which they found recorded in the writings of astronomers. Some of the statements left by ancient writers are indeed ludicrously incongruous; insomuch that Grotius not unjustly said of the account of the constellations given by the poet Aratus, that it could be assigned to no fixed epoch and to no fixed place. However, this would not be the place to discuss details such as are involved in exact inquiries. I have indicated some of these in an appendix to my treatise on 'Saturn,' and others in the preface to my 'Gnomonic Star Atlas'; but for the most part they do not admit very readily of familiar description. Let us turn to less technical considerations, which fortunately are in this case fully as much to the point as exact inquiries, seeing that there is no real foundation for such inquiries in any of the available evidence.

The first obvious feature of the old constellations is one which somehow has not received the attention it deserves. It is as instructive as any of those which have been made the subject of profound research.

There is a great space in the heavens over which none of the old constellations extend, except the River Eridanus as now pictured, but we do not know where this winding stream of stars was supposed by the old observers to come to an end. This great space surrounds the southern pole of the heavens, and thus shows that the first observers of the stars were not acquainted with the constellations which can be seen only from places far south of Chaldaea, Persia, Egypt, India, China, and indeed of all the regions to which the invention of astronomy has been assigned. Whatever the first astronomers were, however profound their knowledge of astronomy may have been (as some imagine), they had certainly not travelled far enough towards the south to know the constellations around the southern pole. If they had been as well acquainted with geography as some assert, if even any astronomer had travelled as far south as the equator, we should certainly have had pictured in the old star charts some constellations in that region of the heavens wherein modern astronomers have placed the Octant, the Bird of Paradise, the Sword-fish, the Flying-fish, Toucan, the Net, and other uncelestial objects.

In passing I may note that this fact disposes most completely of a theory lately advanced that the constellations were invented in the southern hemisphere, and that thus is to be explained the ancient tradition that the sun and stars have changed their courses. For though all the northern constellations would have been more or less visible from parts of the southern hemisphere near the equator, it is absurd to suppose that a southern observer would leave untenanted a full fourth of the heavens round the southern or visible pole, while carefully filling up the space around the northern or unseen pole with incomplete constellations whose northern unknown portions would include that pole. Supposing it for a moment to be true, as a modern advocate of the southern theory remarks, that 'one of the race migrating from one side to the other of the equator would take his position from the sun, and fancy he was facing the same way when he looked at it at noon, and so would think the motion of the stars to have altered instead of his having turned round,' the theory that astronomy was brought to us from south of the equator cannot possibly be admitted in presence of that enormous vacant region around the southern pole. I think, however, that, apart from this, a race so profoundly ignorant as to suppose any such thing, to imagine they were looking north when in reality they were looking south, can hardly be regarded as the first founders of the science of astronomy.

The great gap I have spoken of has long been recognised. But one remarkable feature in its position has not, to the best of my remembrance, been considered—the vacant space is eccentric with regard to the southern pole of the heavens. The old constellations, the Altar, the Centaur, and the ship Argo, extend within twenty degrees of the pole, while the Southern Fish and the great sea-monster Cetus, which are the southernmost constellations on the other side, do not reach within some sixty degrees of the pole.

Of course, in saying that this peculiarity has not been considered, I am not suggesting that it has not been noticed, or that its cause is in any way doubtful or unknown. We know that the earth, besides whirling once a day on its axis, and rushing on its mighty orbit around the sun (spanning some 184,000,000 of miles) reels like a gigantic top, with a motion so slow that 25,868 years are required for a single circuit of the swaying axis around an imaginary line upright to the plane in which the earth travels. And we know that in consequence of this reeling motion the points of the heavens opposite the earth's poles necessarily change. So that the southern pole, now eccentrically placed amid the region where there were no constellations in old times, was once differently situated. But the circumstance which seems to have been overlooked is this, that by calculating backwards to the time when the southern pole was in the centre of that vacant region, we have a much better chance of finding the date (let us rather say the century) when the older constellations were formed, than by any other process. We may be sure not to be led very far astray; for we are not guided by one constellation but by several, whereas all the other indications which have been followed depend on the supposed ancient position of single constellations. And then most of the other indications are such as might very well have belonged to periods following long after the invention of the constellations themselves. An astronomer might have ascertained, for instance, that the sun in spring was in some particular part of the Ram or of the Fishes, and later a poet like Aratus might describe that relation (erroneously for his own epoch) as characteristic of one or other constellation; but who is to assure us that the astronomer who noted the relation correctly may not have made his observation many hundreds of years after those constellations were invented? Whereas, there was one period, and only one period, when the most southernmost of the old constellations could have marked the limits of the region of sky visible from some northern region. Thus, too, may we form some idea of the latitude in which the first observers lived. For in high latitudes the southernmost of the old constellations would not have been visible at all, and in latitudes much lower than a certain latitude, presently to be noted, these constellations would have ridden high above the southern horizon, other star-groups showing below them which were not included among the old constellations.

I have before me as I write a picture of the southern heavens, drawn by myself, in which this vacant space—eccentric in position but circular in shape—is shown. The centre lies close by the Lesser Magellanic cloud—between the stars Kappa Toucani and Eta Hydri of our modern maps, but much nearer to the last named. Near this spot, then, we may be sure, lay the southern pole of the star-sphere when the old constellations, or at least the southern ones, were invented. (If there had been astronomers in the southern hemisphere Eta Hydri would certainly have been their pole-star.)

Now it is a matter of no difficulty whatever to determine the epoch when the southern pole of the heavens was thus placed.[57] Between 2100 and 2200 years before the Christian era the southern constellations had the position described, the invisible southern pole lying at the centre of the vacant space of the star-sphere—or rather of the space free from constellations. It is noteworthy that for other reasons this period, or rather a definite epoch within it, is indicated as that to which must be referred the beginning of exact astronomy. Amongst others must be mentioned this—that in the year 2170 B.C. quam proxime, the Pleiades rose to their highest above the horizon at noon (or technically made their noon culmination), at the spring equinox. We can readily understand that to minds possessed with full faith in the influence of the stars on the earth, this fact would have great significance. The changes which are brought about at that season of the year, in reality, of course, because of the gradual increase in the effect of the sun's rays as he rises higher and higher above the celestial equator, would be attributed, in part at least, to the remarkable star-cluster coming then close by the sun on the heavens, though unseen. Thus we can readily understand the reference in Job to the 'sweet influences of the Pleiades.' Again at that same time, 2170 B.C. when the sun and the Pleiades opened the year (with commencing spring) together, the star Alpha of the Dragon, which was the pole-star of the period, had that precise position with respect to the true pole of the heavens which is indicated by the slope of the long passage extending downwards aslant from the northern face of the Great Pyramid; that is to say, when due north below the pole (or at what is technically called its sub-polar meridional passage) the pole-star of the period shone directly down that long passage, and I doubt not could be seen not only when it came to that position during the night, but also when it came there during the day-time.

But some other singular relations are to be noted in connection with the particular epoch I have indicated.

It is tolerably clear that in imagining figures of certain objects in the heavens, the early observers would not be apt to picture these objects in unusual positions. A group of stars may form a figure so closely resembling that of a familiar object that even a wrong position would not prevent the resemblance from being noticed, as for instance the 'Chair,' the 'Plough,' and so forth. But such cases are not numerous; indeed, to say the truth, one must 'make believe a good deal' to see resemblance between the star-groups and most of the constellation-figures, even under the most favourable conditions. When there is no very close resemblance, as is the case with all the large constellations, position must have counted for something in determining the association between a star-group and a known object.

Now the constellations north of the equator assume so many and such various positions that this special consideration does not apply very forcibly to them. But those south of the equator are only seen above the southern horizon, and change little in position during their progress from east to west of the south point. The lower down they are the less they change in position. And the very lowest—such as those were, for instance, which I have been considering in determining the position of the southern pole—are only fully visible when due south. They must, then, in all probability, have stood upright or in their natural position when so placed, for if they were not rightly placed then they only were so when below the horizon and consequently invisible.

Let us, then, inquire what was the position of the southernmost constellations when fully seen above the southern horizon at midnight.

The Centaur stood then as he does now, upright; only—whereas now in Egypt, Chaldaea, India, Persia, and China, only the upper portions of his figure rise above the horizon, he then stood, the noblest save Orion of all the constellations, with his feet (marked by the bright Alpha and Beta still belonging to the constellation, and by the stars of the Southern Cross which have been taken from it) upon the horizon itself. In latitude twenty degrees or so north he may still be seen thus placed when due south.

The Centaur was represented in old times as placing an offering upon the altar, which was pictured, says Manilius, as bearing a fire of incense represented by stars. This to a student of our modern charts seems altogether perplexing. The Centaur carries the wolf on the end of his spear; but instead of placing the wolf (not a very acceptable meat offering, one would suppose) upon the altar, he is directing this animal towards the base of the altar, whose top is downwards, the flames represented there tending (naturally) downwards also. It is quite certain the ancient observers did not imagine anything of this sort. As I have said, Aratus tells us the celestial Centaur was placing an offering upon the altar, which was therefore upright, and Manilius describes the altar as

Ferens thuris, stellis imitantibus, ignem,

so that the fire was where it should be, on the top of an upright altar, where also on the sky itself were stars looking like the smoke from incense fires. Now that was precisely the appearance presented by the stars forming the constellation at the time I have indicated, some 2170 years B.C. Setting the altar upright above the southern horizon (that is, inverting the absurd picture at present given of it) we see it just where it should be placed to receive the Centaur's offering. A most remarkable portion of the Milky Way is then seen to be directly above the altar in such a way as to form a very good imitation of smoke ascending from it. This part of the Milky Way is described by Sir J. Herschel, who studied it carefully during his stay at the Cape of Good Hope, as forming a complicated system of interlaced streaks and masses which covers the tail of Scorpio (extending from the altar which lies immediately south of the Scorpion's Tail). The Milky Way divides, in fact, just above the altar as the constellation was seen 4000 years ago above the southern horizon, one branch being that just described, the other (like another stream of smoke) 'passing,' says Herschel, 'over the stars Iota of the Altar, Theta and Iota of the Scorpion, etc., to Gamma of the Archer, where it suddenly collects into a vivid oval mass, so very rich in stars that a very moderate calculation makes their number exceed 100,000.' Nothing could accord better with the descriptions of Aratus and Manilius.

But there is another constellation which shows in a more marked way than either the Centaur or the Altar that the date when the constellations were invented must have been near that which I have named. Both Ara and Centaurus look now in suitable latitudes (about twenty degrees north) as they looked in higher latitudes (about forty degrees north) 4000 years ago. For, the reeling motion of our earth has changed the place of the celestial pole in such a way as only to depress these constellations southwards without much changing their position; they are nearly upright when due south now as they were 4000 years ago, only lower down. But the great ship Argo has suffered a much more serious displacement. One cannot now see this ship like a ship at any time or from any place on the earth's surface. If we travel south till the whole constellation comes into visibility above the southern horizon at the proper season (January and February for the midnight hours) the keel of the ship is aslant, the stern being high above the waist (the fore part is wanting). If we travel still further south, we can indeed reach places where the course of the ship is so widened, and the changes of position so increased, that she appears along part of her journey on an even keel, but then she is high above the horizon. Now 4000 years ago she stood on the horizon itself at her southern culmination, with level keel and upright mast.

In passing I may note that for my own part I imagine that this great ship represented the Ark, its fore part being originally the portion of the Centaur now forming the horse, so that the Centaur was represented as a man (not as a man-horse) offering a gift on the Altar. Thus in this group of constellations I recognise the Ark, and Noah going up from the Ark towards the altar 'which he builded unto the Lord; and took of every clean beast, and of every clean fowl, and offered burnt offerings on the altar.' I consider further that the constellation-figures of the Ship, the Man with an offering, and the Altar, painted or sculptured in some ancient astrological temple, came at a later time to be understood as picturing a certain series of events, interpreted and expanded by a poetical writer into a complete narrative. Without venturing to insist on so heterodox a notion, I may remark as an odd coincidence that probably such a picture or sculpture would have shown the smoke ascending from the Altar which I have already described, and in this smoke there would be shown the bow of Sagittarius; which, interpreted and expanded in the way I have mentioned, might have accounted for the 'bow set in the clouds, for a token of a covenant.' It is noteworthy that all the remaining constellations forming the southern limit of the old star-domes or charts, were watery ones—the Southern Fish, over which Aquarius is pouring a quite unnecessary stream of water, the Great Sea Monster towards which in turn flow the streams of the River Eridanus. The equator, too, was then occupied along a great part of its length by the great sea serpent Hydra, which reared its head above the equator, very probably indicated then by a water horizon, for nearly all the signs below it were then watery. At any rate, as the length of Hydra then lay horizontally above the Ship, whose masts reached it, we may well believe that this part of the picture of the heavens showed a sea-horizon and a ship, the great sea serpent lying along the horizon. On the back of Hydra is the Raven, which again may be supposed by those who accept the theory mentioned above to have suggested the raven which went forth to and fro from the ark. He is close enough to the rigging of Argo to make an easy journey of it. The dove, however, must not be confounded with the modern constellation Columba, though this is placed (suitably enough) near the Ark. We must suppose the idea of the dove was suggested by a bird pictured in the rigging of the celestial ship. The sequence in which the constellations came above the horizon as the year went round corresponded very satisfactorily with the theory, fanciful though this seem to some. First Aquarius pouring streams of water, the three fishes (Pisces and Piscis australis), and the great sea monster Cetus, showing how the waters prevailed over the highest hills, then the Ark sailing on the waters, a little later the Raven (Corvus), the man descending from the ark and offering a gift on the Altar, and last the Bow set amid the clouds.

The theory just described may not meet with much favour. But wilder theories of the story of the deluge have been adopted and advocated with considerable confidence. One of the wildest, I fear, is the Astronomer-Royal's, that the deluge was simply a great rising of the Nile; and Sir G. Airy is so confident respecting this that he says, 'I cannot entertain the smallest doubt that the flood of Noah was a flood of the Nile;' precisely as he might say, 'I cannot entertain the smallest doubt that the earth moves round the sun.' On one point we can entertain very little doubt indeed. If it ever rained before the flood, which seems probable, and if the sun ever shone on falling rain, which again seems likely, nothing short of a miracle could have prevented the rainbow from making its appearance before the flood. The wildest theory that can be invented to explain the story of the deluge cannot be wilder than the supposition that the rays of sunlight shining on falling raindrops could have ever failed to show the prismatic colours. The theory I have suggested above, without going so far as strongly to advocate it, far less insist upon it, is free at any rate from objection on this particular score, which cannot be said of the ordinary theory. I am not yet able, however, to say that 'I cannot entertain the smallest doubt' about my theory.

We may feel tolerably sure that the period when the old southern constellations were formed must have been between 2400 and 2000 years before the present era, a period, by the way, including the date usually assigned to the deluge,—which, however, must really occupy our attention no further. In fact, let us leave the watery constellations lying below the equator of those remote times and seek at once the highest heavens above them.

Here, at the northern pole of these days, we find the great Dragon, which in any astrological temple of the time must have formed the highest or crowning constellation, surrounding the very key-stone of the dome. He has fallen away from that proud position since. In fact, even 4000 years ago he only held to the pole, so to speak, by his tail, and we have to travel back 2000 years or so to find the pole situate in a portion of the length of the Dragon which can be regarded as central. One might almost, if fancifully disposed, recognise the gradual displacement of the Dragon from his old place of honour, in certain traditions of the downfall of the great Dragon whose 'tail drew the third part of the stars of heaven.'

The central position of the Dragon, for even when the pole-star had drawn near to the Dragon's tail the constellation was still central, will remind the classical reader of Homer's description of the Shield of Hercules—

The scaly horror of a dragon, coil'd Full in the central field, unspeakable, With eyes oblique retorted, that ascant Shot gleaming fire. (Elton's translation.)

I say Homer's description, for I cannot understand how any one who compares together the description of the Shield of Achilles in the Iliad and that of the Shield of Hercules in the fragmentary form in which we have it, can doubt for a moment that both descriptions came from the same hand. (The theory that Hesiod composed the latter poem can scarcely be entertained by any scholar.) As I long since pointed out in my essay 'A New Theory of Achilles' Shield' ('Light Science,' first series), no poet so inferior as actually to borrow Homer's words in part of the description of the Shield of Hercules could have written the other parts not found in the Shield of Achilles. 'I cannot for my own part entertain the slightest doubt'—that is to say, I think it altogether probable—that Homer composed the lines supposed to describe the Shield of Hercules long before he introduced the description, pruned and strengthened, into that particular part of the Iliad where it served his purpose best. And I have as little doubt that the original description, of which we only get fragments in either poem, related to something far more important than a shield. The constellations are not suitable adornments for the shield of fighting man, even though he was under the special care of a celestial mother and had armour made for him by a celestial smith. Yet we learn that Achilles' shield displayed—

The starry lights that heav'n's high convex crown'd The Pleiads, Hyads, and the northern beam, And great Orion's more refulgent beam,— To which, around the cycle of the sky, The bear revolving, points his golden eye,— Still shines exalted.

And so forth. The Shield of Hercules displayed at its centre the polar constellation the Dragon. We read also that—

There was the knight of fair-hair'd Danae born, Perseus.

Orion is not specially mentioned, but Orion, Lepus, and the Dogs seem referred to:—

Men of chase Were taking the fleet hares; two keen-toothed dogs Bounded beside.

Homer would find no difficulty in pluralising the mighty Hunter and the hare into huntsmen and hares when utilising a description originally referring to the constellation.

I conceive that the original description related to one of those zodiac temples whose remains are still found in Egypt, though the Egyptian temples of this kind were probably only copies of more ancient Chaldaean temples. We know from Assyrian sculptures that representations of the constellations (and especially the zodiacal constellations) were common among the Babylonians; and, as I point out in the essay above referred to, 'it seems probable that in a country where Sabaeanism or star-worship was the prevailing form of religion, yet more imposing proportions would be given to zodiac temples than in Egypt.' My theory, then, respecting the two famous 'Shields' is that Homer in his eastern travels visited imposing temples devoted to astronomical observation and star-worship, and that nearly every line in both descriptions is borrowed from a poem in which he described a temple of this sort, its domed zodiac, and those illustrations of the labours of different seasons and of military or judicial procedures which the astrological proclivities of star-worshippers led them to associate with the different constellations. For the arguments on which this theory is based I have not here space. They are dealt with in the essay from which I have quoted.

One point only I need touch upon here, besides those I have mentioned already. It may be objected that the description of a zodiac temple has nothing to connect it with the subject of the Iliad. This is certainly true; but no one who is familiar with Homer's manner can doubt that he would work in, if he saw the opportunity, a poem on some subject outside that of the Iliad, so modifying the language that the description would correspond with the subject in hand. There are many passages, though none of such length, in both the Iliad and the Odyssey, which seem thus to have been brought into the poem; and other passages not exactly of this kind yet show that Homer was not insensible to the advantage of occasionally using memory instead of invention.

Any one who considers attentively the aspect of the constellation Draco in the heavens, will perceive that the drawing of the head in the maps is not correct; the head is no longer pictured as it must have been conceived by those who first formed the constellation. The two bright stars Beta and Gamma are now placed on a head in profile. Formerly they marked the two eyes. I would not lay stress on the description of the Dragon in the Shield of Hercules, 'with eyes oblique retorted, that askant shot gleaming fire;' for all readers may not be prepared to accept my opinion that that description related to the constellation Draco. But the description of the constellation itself by Aratus suffices to show that the two bright stars I have named marked the eyes of the imagined monster—in fact, Aratus's account singularly resembles that given in the Shield of Hercules. 'Swol'n is his neck,' says Aratus of the Dragon—

... Eyes charg'd with sparkling fire His crested head illume. As if in ire, To Helice he turns his foaming jaw, And darts his tongue, barb'd with a blazing star.

And the dragon's head with sparkling eyes can be recognised to this day, so soon as this change is made in its configuration, whereas no one can recognise the remotest resemblance to a dragon's head in profile. The star barbing the Dragon's tongue would be Xi of the Dragon according to Aratus's account, for so only would the eyes be turned towards Helice the Bear. But when Aratus wrote, the practice of separating the constellations from each other had been adopted; in fact, he derived his knowledge of them chiefly from Eudoxus, the astronomer and mathematician, who certainly would not have allowed the constellations to be intermixed. In the beginning, there are reasons for believing it was different, and if a group of stars resembled any known object it would be called after that object, even though some of the stars necessary to make up the figure belonged already to some other figure. This being remembered, we can have no difficulty in retorting the Dragon's head more naturally—not to the star Xi of the Dragon, but to the star Iota of Hercules. The four stars are situated thus, the larger ones representing the eyes; and so far as the head is concerned it is a matter of indifference whether the lower or the upper small star be taken to represent the tongue. But, as any one will see who looks at these stars when the Dragon is best placed for ordinary (non-telescopic) observation, the attitude of the animal is far more natural when the star Iota of Hercules marks the tongue, for then the creature is situated like a winged serpent hovering above the horizon and looking downwards, whereas when the star Xi marks the tongue, the hovering Dragon is looking upwards and is in an unnaturally constrained position. (I would not, indeed, claim to understand perfectly all the ways of dragons; still it may be assumed that a dragon hovering above the horizon would rather look downwards in a natural position than upwards in an awkward one.)

The star Iota of Hercules marks the heel of this giant, called the Kneeler (Engonasin) from time immemorial. He must have been an important figure on the old zodiac temples, and not improbably his presence there as one of the largest and highest of the human figures may have caused a zodiac-dome to be named after Hercules. The Dome of Hercules would come near enough to the title, 'The Shield of Hercules,' borne by the fragmentary poem dealt with above. The foot of the kneeling man was represented on the head of the dragon, the dragon having hold of the heel. And here, again, some imagine that a sculptured representation of these imagined figures in the heavens may have been interpreted and expanded into the narrative of a contest between the man and the old serpent the dragon, Ophiuchus the serpent-bearer being supposed to typify the eventual defeat of the dragon. This fancy might be followed out like that relating to the deluge; but the present place would be unsuitable for further inquiries in that particular direction.

Some interest attaches to the constellation Ophiuchus, to my mind, in the evidence it affords respecting the way in which the constellations were at first intermixed. I have mentioned one instance in which, as I think, the later astronomers separated two constellations which had once been conjoined. Many others can be recognised when we compare the actual star-groups with the constellation-figures as at present depicted. No one can recognise the poop of a ship in the group of stars now assigned to the stern of Argo, but if we include the stars of the Greater Dog, and others close by, a well-shaped poop can be clearly seen. The head of the Lion of our maps is as the head of a dog, so far as stars are concerned; but if stars from the Crab on one side and from Virgo on the other be included in the figure, and especially Berenice's hair to form the tuft of the lion's tail, a very fine lion with waving mane can be discerned, with a slight effort of the imagination. So with Bootes the herdsman. He was of old 'a fine figure of a man,' waving aloft his arms, and, as his name implies, shouting lustily at the retreating bear. Now, and from some time certainly preceding that of Eudoxus, one arm has been lopped off to fashion the northern crown, and the herdsman holds his club as close to his side as a soldier holds his shouldered musket. The constellation of the Great Bear, once I conceive the only bear (though the lesser bear is a very old constellation), has suffered wofully. Originally it must have been a much larger bear, the stars now forming the tail marking part of the outline of the back; but first some folks who were unacquainted with the nature of bears turned the three stars (the horses of the plough) into a long tail, abstracting from the animal all the corresponding portion of his body, and then modern astronomers finding a great vacant space where formerly the bear's large frame extended, incontinently formed the stars of this space into a new constellation, the Hunting Dogs. No one can recognise a bear in the constellation as at present shaped, but any one who looks attentively at the part of the skies occupied by the constellation will recognise (always 'making believe a good deal') a monstrous bear, with the proper small head of creatures of the bear family, and with exceedingly well-developed plantigrade feet. Of course this figure cannot at all times be recognised with equal facility; but before midnight during the last four or five months in the year, the bear occupies positions favouring his recognition, being either upright on his feet, or as if descending a slope, or squatting on his great haunches. As a long-tailed animal the creature is more like one of those wooden toy-monkeys which used to be made for children, and may be now, in which the sliding motion of a ringed rod carried the monkey over the top of a stick. The little bear has I think been borrowed from the dragon, which was certainly a winged monster originally.

Now the astronomers who separated from each other, and in so doing spoiled the old constellation-figures, seem to have despaired of freeing Ophiuchus from his entanglements. The Serpent is twined around his body, the Scorpion is clawing at one leg. The constellation makers have per fas et nefas separated Scorpio from the Serpent Holder, spoiling both figures. But the Serpent has been too much for them, insomuch that they have been reduced to the abject necessity of leaving one part of the Serpent on one side of the region they allow to Ophiuchus, and the other part of the Serpent to the other.

A group of constellations whose origin and meaning are little understood remains to be mentioned. Close by the Dragon is King Cepheus, beside him his wife Cassiopeia (the Seated Lady), near whom is Andromeda the Chained Lady. The Sea Monster Cetus is not far away, though not near enough to threaten her safety, the Ram and Triangle being between the monster's head and her feet, the Fishes intervening between the body of the monster and her fair form. Close at hand is Perseus, the Rescuer, with a sword (looking very much like a reaping-hook in all the old pictures) in his right hand, and bearing in his left the head of Medusa. The general way of accounting for the figures thus associated has been by supposing that, having a certain tradition about Cepheus and his family, men imagined in the heavens the pictorial representation of the events of the tradition. I have long believed that the actual order in this and other cases was the reverse of this, that men imagined certain figures in the heavens, pictured these figures in their astronomical temples or observatories, and made stories to fit the pictures afterwards, probably many generations afterwards. Be this as it may, we can at present give no satisfactory explanation of the group of constellations.

Wilford gives an account, in his 'Asiatic Researches,' of a conversation with a pundit or astronomer respecting the names of the Indian constellations. 'Asking him,' he says, 'to show me in the heavens the constellation Antarmada, he immediately pointed to Andromeda, though I had not given him any information about it beforehand. He afterwards brought me a very rare and curious work in Sanscrit, which contained a chapter devoted to Upanachatras, or extra-zodiacal constellations, with drawings of Capuja (Cepheus) and of Casyapi (Cassiopeia) seated and holding a lotus-flower in her hand, of Antarmada charmed with the Fish beside her, and last of Paraseia (Perseus), who, according to the explanation of the book, held the head of a monster which he had slain in combat; blood was dropping from it, and for hair it had snakes.' Some have inferred from the circumstance that the Indian charts thus showed the Cassiopeian set of constellations, that the origin of these figures is to be sought in India. But probably both the Indian and the Greek constellation-figures were derived from a much older source.

The zodiacal twelve are in some respects the most important and interesting of all the ancient constellations. If we could determine the origin of these figures, their exact configuration as at first devised, and the precise influences assigned to them in the old astrological systems, we should have obtained important evidence as to the origin of astronomy itself. Not indeed that the twelve signs of the zodiac were formed at the beginning or even in the early infancy of astronomy. It seems abundantly clear that the division of the zodiac (which includes the moon's track as well as the sun's) had reference originally to the moon's motions. She circuits the star-sphere in about twenty-seven days and a third, while the lunation or interval from new moon to new moon is, as we all know, about twenty-nine days and a half in length. It would appear that the earliest astronomers, who were of course astrologers also, of all nations—the Indian, Egyptian, Chinese, Persian, and Chaldaean astronomers—adopted twenty-eight days (probably as a rough mean between the two periods just named) for their chief lunar period, and divided the moon's track round the ecliptic into twenty-eight portions or mansions. How they managed about the fractions of days outstanding—whether the common lunation was considered or the moon's motion round the star-sphere—is not known. The very circumstance, however, that they were for a long time content with their twenty-eight lunar mansions shows that they did not seek great precision at first. Doubtless they employed some rough system of 'leap-months' by which, as occasion required, the progress of the month was reconciled with the progress of the moon, just as by our leap-years the progress of the year is reconciled with the progress of the sun or seasons.

The use of the twenty-eight-day period naturally suggested the division of time into weeks of seven days each. The ordinary lunar month is divided in a very obvious manner into four equal parts by the lunar aspects. Every one can recognise roughly the time of full moon and the times of half moon before and after full, while the time of new moon is recognised from these two last epochs. Thus the four quarters of the month, or roughly the four weeks of the month, would be the first time-measure thought of;—after the day, which is the necessary foundation of all time measures. The nearest approach which can be made to a quarter-month in days is the week of seven days; and although some little awkwardness arose from the fact that four weeks differ appreciably from a lunar month, this would not long prevent the adoption of the week as a measure of time. In fact, just as our years begin on different days of the week without causing any inconvenience, so the ancient months might be made to begin with different week-days. All that would be necessary to make the week measure fairly well the quarters of the month, would be to start each month on the proper or nearest week-day. To inform people about this, some ceremony could be appointed for the day of the new moon, and some signal employed to indicate the time when this ceremony was to take place. This—the natural and obvious course—we find was the means actually adopted, the festival of the new moon and the blowing of trumpets in the new moon being an essential part of the arrangements adopted by nations who used the week as a chief measure of time. The seven days were not affected by the new moons so far as the nomenclature of these days, or special duties connected with any one of them, might be concerned.

Originally the idea may have been to have festivals and sacrifices at the time of new moon, first quarter, full moon, and third quarter; but this arrangement would naturally (and did, as we know, actually) give way before long to a new moon festival regulating the month and seventh-day festivals, each class of festival having its appropriate sacrifices and duties. This, I say, was the natural course. Its adoption may have been aided by the recognition of the fact that the seven planets of the old system of astronomy might conveniently be taken to rule the days and the hours in the way described in the essay on astrology. That that nomenclature and that system of association between the planets and the hours, days, and weeks of time-measurement was eventually adopted, is certain; but whether the convenience and apparent mystical fitness of this arrangement led to the use of weekly festivals in conjunction with monthly ones, or whether those weekly festivals were first adopted in the way described above, or whether (which seems altogether more likely) both sets of considerations led to the arrangement, we cannot certainly tell. The arrangement was in every way a natural one; and one may say, considering all the circumstances, that it was almost an inevitable one.

There was, however, another possible arrangement, viz., the division of time into ten-day periods, three to each month, with corresponding new moon festivals. But as the arrival of the moon at the thirds of her progress are not at all so well marked as her arrival at the quarters, and as there is no connection between the number ten and the planets, this arrangement was far less likely to be adopted than the other. Accordingly we find that only one or two nations adopted it. Six sets of five days would be practically the same arrangement; five sets of six for each month would scarcely be thought of, as with that division the use of simple direct observations of the moon for time measurement, which was the real aim of all such divisions, would not be convenient or indeed even possible for the generality of persons. Few could tell easily when the moon is two-fifths or four-fifths full, whereas every one can tell when she is half-full or quite full (the requisite for weekly measurement); and it would be possible to guess pretty nearly when she is one-third or two-thirds full, the requisite for the tridecennial division.

My object in the above discussion of the origin of the week (as distinguished from the origin of the Sabbath, which I considered in the essay on astrology), has been to show that the use of the twelve zodiacal signs was in every case preceded by the use of the twenty-eight lunar mansions. It has been supposed that those nations in whose astronomy the twenty-eight mansions still appear, adopted one system, while the use of the twelve signs implies that another system had been adopted. Thus the following passage occurs in Mr. Blake's version of Flammarion's 'History of the Heavens:'—'the Chinese have twenty-eight constellations, though the word sion does not mean a group of stars, but simply a mansion or hotel. In the Coptic and ancient Egyptian the word for constellations has the same meaning. They also have twenty-eight, and the same number is found among the Arabians, Persians, and Indians. Among the Chaldaeans or Accadians we find no sign of the number twenty-eight. The ecliptic, or "yoke of the sky," with them, as we see in the newly-discovered tablet, was divided into twelve divisions, as now, and the only connection that can be imagined between this and the twenty-eight is the opinion of M. Biot, who thinks that the Chinese had originally only twenty-four mansions, four more being added by Chenkung, 1100 B.C., and that they corresponded with the twenty-four stars, twelve to the north and twelve to the south, that marked the twelve signs of the zodiac amongst the Chaldaeans. But under this supposition the twenty-eight has no reference to the moon, whereas we have every reason to believe it has.' The last observation is undoubtedly correct—the twenty-eight mansions have been mansions of the moon from the beginning. But in this very circumstance, as also in the very tablets referred to in the preceding passage, we find all the evidence needed to show that originally the Chaldaeans divided the zodiac into twenty-eight parts. For we find from the tablets that, like the other nations who had twenty-eight zodiacal mansions, the Chaldaeans used a seven-day period, derived from the moon's motions, every seventh day being called sabbatu, and held as a day of rest. We may safely infer that the Chaldaean astronomers, advancing beyond those of other nations, recognised the necessity of dividing the zodiac with reference to the sun's motions instead of the moon's. They therefore discarded the twenty-eight lunar mansions, and adopted instead twelve solar signs; this number twelve, like the number twenty-eight itself, being selected merely as the most convenient approximation to the number of parts into which the zodiac was naturally divided by another period. Thus the twenty-eighth part of the zodiac corresponds roughly with the moon's daily motion, and the twelfth part of the zodiac corresponds roughly with the moon's monthly motion; and both the numbers twenty-eight and twelve admit of being subdivided, while twenty-nine (a nearer approach than twenty-eight to the number of days in a lunation) and thirteen (almost as near an approach as twelve to the number of months in a year) do not.