i Cornell lllnivetstt^ OF THE mew J^orK State College of Hfiriculture ..^i^.,..£r.SrSA --. A.S.)-te-U°- Cornell University Library QB 51.P982 Other suns than ours.A series of essays 3 1924 002 960 247 Cornell University Library The original of tliis bool< is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924002960247 OTHER SUNS THAN OURS z o u o o 'A •A W ■^ =^ ■a^'ebula "■ less. , (PECASUSJ ■—♦41 ♦W Fig. 2.— Key to the Map of Many Stars given on opposite side. \Tofacejiage 1. OTHER SUNS THAN OURS. CHAPTER I. THE NEW STAR IN THE ANDROMEDA NEBULA. Although the new star which has appeared in the Constellation Andromeda, in the very heart of the great Andromeda Nebula, is not in itself so remark- able an object as many of the new stars recorded in past years, it throws a stronger and clearer light than any of those orbs on the architecture of the heavens. It answers questions which have been asked for more than two centuries, nay, almost from the very year when the telescope was invented. It discloses much of the real nature of the galaxy — our " island universe," as Humboldt poetically called the stellar system ; and though it is silent respecting the system of galaxies in which many astronomers have believed, it is more eloquent in silence than all the stars in the heavens have been in direct testimony. I propose to consider the aspect in which we must hereafter view the galaxy, noting what had before been imagined, suggested, or proved, and how the new star illuminates the whole system of celestial architecture. But first it is essential that the true position of the new star should be recognized. I cannot hope for attention to the subject on which a new star in the Andromeda Nebula would assuredly 2 OTHER SUNS THAN OURS. throw light, if it remains doubtful whether the orb which appears to be in the heart of that nebula is really there or not. If the new star be not physically associated with the Andromeda Nebula, then it is only interesting as a new star of a somewhat novel order (as will presently be seen), not at all as throw- ing new light on the structure of the stellar universe. It might, for aught that its position actually proves, be many millions of miles nearer than the nebula, or many millions of miles farther away. In either case it would merit only such attention as astronomers gave to the new star in the Northern Crown in 1866, and to the less brilliant star which appeared ten years later in the Swan. It will, of course, be admitted that the apparent agreement of the new star in position with the rich region of nebulous light within the large Andromeda Nebula, affords in itself a strong argument from pro- bability that the star and the nebula are physically associated. It may be said that, so far as we know, a new star may appear anywhere, and that there can be no reason why the particular point on the star sphere where a new star appears should not be a point already occupied by nebulous light. So it might be said that since, as among all possible deals at whist, any particular arrangement is as likely to appear as any other, we ought not to be surprised if a particular deal gives all the trumps to the dealer, and to each of the other players ail the thirteen cards of a single suit. Yet, as a matter of fact, we know that such an arrangement would surprise every one present ; and would incline each to believe that before the deal the cards had been specially arranged. Nothing but the clearest evidence that there had been no special " manipulation " either before or during the deal would convince the observers that what they had seen had resulted from mere chance-coincidence. Their thought would be sound and just. As a THE NEW STAR. 3 problem in what is called inverse probability, an ex- planation which would account for the observed strange fact without requiring that one chance among many thousands of millions had come off would antecedently be far more likely than one which involved so strange a coincidence. So, assuredly, it is in the case of the new star in the Andromeda Nebula. We cannot but be surprised — regarding the matter at first as depending on chance- coincidence — that so unusual an object as a new star should make its appearance in the midst of the richest part of the light from a nebula so remarkable that it has been called by astronomers in former times " the transcendently beautiful queen of the nebulae." In any nebula a new star would be strange, insomuch that few astronomers would feel any doubt about the existence of some connection between nebula and star. But a new star in the most won- derful and beautiful of all star cloudlets ! — certainly that would be a strange chance, if it could be shown really to be casual. Nothing but the clearest evidence could convince the astronomer that this was the true explanation — though it must be admitted also that nothing but clear evidence of physical associa- tion, outside the evidence derived from' position, would justify any astronomer in saying that such a physic£il association was a demonstrated fact. We naturally turn to other new stars to see if they have given any evidence bearing on the peculiar position of the new star in Andromeda. If, for example, we find that new stars, or variable stars (for most new stars are only stars which vary greatly in lustre) affect nebulae, we shall be prepared to admit that the evidence of physical association be- tween the new star and its nebular surroundings is greatly strengthened. The only new stars we need consider are those which have appeared in recent times, since it is only B 2 4 OTHER SUNS THAN OURS. within recent times that the greater number of the nebulae have been discovered. The star Eta Argus must be regarded as one of those varying sufficiently in splendour to be within our category. This star ranges in brightness from bare visibility to a lustre exceeding that of every star in the heavens except Sirius alone. In 1 840 it was shining with this maximum splendour. It had risen to its greatest glory by a succession of leaps, with occasional retrogressions. After 1840, Eta rapidly diminished in brightness, then faded more gradually, until it reached its present lustre, that of a sixth magnitude star. It is manifest that if we were at a greater distance from this strange variable, it would be now invisible, and therefore its return to the splendour of 1840 would correspond with the appearance of a new star in the heavens. This star, then, may throw light on the question whether nebulae and new stars are apt to be associated. The apparent association between Eta Argils and a nebula is of the most striking kind. Eta is in the very brightest part of the most remarkable nebula in the southern hemisphere — the far-reaching mass of star- mist known as the Keyhole Nebula (because of the peculiar shape of a dark space in its midst). It had been held to be possible, but barely possible, that the association between Eta Argils and the sur- rounding nebula was only apparent, not real. It is hardly necessary to point out that this extreme im- probability — that the most remarkable variable star in the southern heavens should be apparently in the very heart of the most remarkable southern nebula, yet not really connected with it, is immensely increased by the appearance of a new star in the midst of the most remarkable northern nebula. The corresponding improbability in the case of the new nebula in An- dromeda does not double the resulting improbability, but introduces an improbability many times greater THE NEW STAR, 5 The chance is one in about a million that, with fair tossing, " heads " will be tossed twenty times run- ning, and, of course, it is extremely improbable that this will happen at a first trial. If it happened, we might reasonably suspect that the tossing was not fair. If it happened at a second trial, the strength of this suspicion would not be doubled, for the chance of twenty heads coming in each of two trials is not one in about two millions (or twice the number in the case of a single trial), but one in about a million millions) — not a million and a million added together, but a million and a million multiplied together. Similarly in the case of new stars and nebulae, the occurrence of two cases where a new or highly variable star is associated with a nebula, increases many thousandfold the probability that such associa- tion is not accidental, but real. Let us, however, consider other new stars. The new star in the Northern Crown was not associated with any nebula. In its case, therefore, we find no evidence confirming, though also (I need hardly say) we find none disproving, or even tending towards disproving, our belief (now almost certainty) that the new star and the nebula in which it has appeared are physitally associated. But the third case — and there are but three — is decisive. In 1876 a new star of the third magnitude appeared in the constellation Cygnus. Like all new stars it gradually lost its newly acquired lustre. But unlike Eta Argfts and the new star in the Northern Crown, the new star in the Swan eventually disappeared — as a star — altogether. Where it had been there was no star, but — there was a nebula ! The nebula was, or rather is, a very faint one— a circular bluish disc, so faint and wan that it would escape notice, had not the new star pointed to it. Doubtless it had escaped notice before the star appeared. For no one supposes 6 OTHER SUNS THAN OURS. that a new body had really come into existence when that so-called new star shone out. Even the regular sky-surveyors, who as a rule are very hard to convince, did not imagine anything quite so unlikely as that The nebula had been there — a faint planetary nebula as now — but it had escaped notice, situated as it is in a singularly crowded region of the heavens. Here, then, was a case where not only had a new star appeared in the midst of a nebula, but where it had made a nebula known to us by so appearing, and had afterwards subsided so completely into the bosom of its parent nebula that no sign of it could any longer be seen. Two, then, out of the only three new stars available for our inquiry have been in the most remarkable manner associated with nebulae. Scarcely one was really needed to convince us that the new star is associated with the remarkable nebula in the midst of which it has appeared ; considering that there are but three new stars to deal with we might very fairly have expected that not one would give us any satis- factory evidence ; but as a matter of fact two out of the three give evidence, and in each case (unlike though the two cases are) it is evidence of the most remarkable kind. To believe in the face of such testimony as this that the new star is not physically associated with the nebula in which it seems to lie, would not be scientific caution but dull and obstinate incredulity. This, however, is by no means all the evidence. Nay, overwhelming though this evidence is, it is by no means the strongest part of the evidence. The new star in the Swan owed the chief part of its lustre — and towards the close all its lustre — to matter of the same nature as the gaseous plane- tary* nebula into which it faded away. It shone at * It may be necessary to explain that this word " planetary ' THE NEW STAR. 7 the last with only three tints, those with which the great Fish-mouth Nebula in Orion, and the great Keyhole Nebula in Argo, are observed to shine. I make no doubt whatever, that if a telescope, armed with a stellar-spectroscope, could have been turned on Eta Argfls in 1840, that resplendent orb would have been found to shine — during its exceptional lustre — with the same kind of light which comes from the nebula around it (besides its own customary stellar light). But evidence of this kind was not then available. If the new star in Andromeda, like the new stars in the Northern Crown and the Swan, showed the signs of gaseous matter, glowing with intense lustre, we should be no wiser as to its physical association with the Andromeda Nebula, than without spectro- scopic evidence : for the Andromeda Nebula is not one of the gaseous kind. The chances were that, formed suddenly ass it had been, the new star would give a spectrum of bright lines indicative of gaseity. This would have decided nothing ; for the star in the Northern Crown, as surely a star as our sun, and now for nearly a score of years shining steadfastly with its original stellar lustre, yet shone for several weeks with light giving a spectrum quite unlike the normal spectrum of a star ; and, which is still more to the purpose, quite unlike the spectrum of that star itself in its normal condition. It was antecedently very unlikely that the new star would show the same spectrum as the nebula in Andromeda ; for the simple reason that not only is this spectrum very unlike the spectrum of any new star yet observed, but it is a very strange, one may relates to the aspect of the nebulae, thus termed. They a/e by no means planetary in the sense of being wandering bodies ; but they present round discs like the planets. Indeed, though very faint and dim, these bodies look much like some very distant planet, only faintly illuminated by the sun's rays. 8 OTHER SUNS THAN OURS. say a unique spectrum. It is not like the spectrum of a star, which is a rainbow-tinted streak crossed by dark lines. It is quite unlike the spectrum of a gaseous nebula, which consists entirely of three or four bright lines. It is not like the spectrum of the new star in the Northern Crown, which was a rain- bow-tinted streak, crossed by dark absorption lines, but also by four intensely bright lines, belonging to the intensely hot gases from which the greater part of that remote sun's light came during its outblazing. Nor is it like the spectrum of the new star in the Swan, at any stage of that star's existence. The spectrum of the Andromeda Nebula is a rainbow-tinted streak, without any bright or dark lines, and only giving evidence of the absorptive action of vapours by a sudden degradation of light near the red end. An unlikely spectrum, one would say, for a new star, even though that new star had its birthplace in the very heart of the Andromeda Nebula. Yet, quite unexpectedly, this peculiar spectrum is just what the new star gives when its light is sifted by the most marvellous of all the instruments of our time — the spectroscope. Carefully observed by spectroscopists of proved skill, and under atmo- spheric conditions so favourable that the delicate light of the gaseous nebulae showed its characteristic spectrum of bright lines, the new star was found to give no bright lines at all, but the same dull con- tinuous rainbow-tinted streak, suddenly degrading at the red end, which is given by the great Andromeda Nebula itself. Even this is not quite all, though this is much more than enough to prove that the new star is physically associated with the nebula. The con- densed part of the Andromeda Nebula had been more than suspected of change during several weeks before the new star appeared, and has been observed THE NEW STAR. 9 to change considerably since that event. The new star itself varies from time to time in aspect. On the 3rd of September it was not so sharply defined as it had been, but had a somewhat hazy aspect ; in technical terms, it was impossible to bring it to a disc. The fault was not in the air, for other stars came sharply into focus. On the following night the new star presented the usual appearance of an eighth magnitude star, neatly and sharply defined. More- over, the new star faded gradually in lustre from the time of its discovery. The new star, then, is in reality, as well as in ap- pearance, in the heart of the great nebula, and we have to consider what, so situated, the strange orb may mean. The new star's teaching, though not new to one or two who had reasoned out from other evidence the same results, is altogether new for those who have retained their belief in the doctrines impressively taught by Humboldt, Arago, Nichol, and others, but never definitely adopted (as they supposed and as is commonly supposed) by Sir William Herschel — the greatest astronomical observer the world has known. In his surveys of the celestial depths. Sir William Herschel recognized about three thousand star- cloudlets of various orders. Sir John Herschel, second only to his father among astronomical ob- servers, completed the survey of the celestial sphere from a station near Cape Town, and added so many more star-cloudlets as to bring up the number to close on five thousand. At present, with all the work of other observers, that is still about the num- ber of known nebulae. Sir W. Herschel suggested many ideas about these objects, or rather many ideas occurred to him in the course of his labours. It may surprise many who know of Herschei's views only through Arago, lO OTHER SUNS THAN OURS. NicholjGuillemin, Flam marion, and others, or through Humboldt, whose knowledge of Herschel's work was derived from imperfect abstracts, to learn that Herschel regarded the rich nebular region in the Constellation Virgo, as consisting of the fragments of a vast galaxy able to extend itself nearer towards us there, because on that side our galaxy has its least extension. It is so commonly believed that Herschel regarded all the stellar nebulae as external galaxies, that this suggestion, according to which hundreds of such nebulae are only fragments of a part of what was once a great galaxy, is worth study- ing, if it were only to show how little is popularly known either of the great master's wonderful work or of his daring conceptions. In 1 8 17 and 1818, when Herschel was very old, at the actual close, indeed, of his observing career, he threw out ideas as to the possible distance of stars and star-cloudlets, if estimated by their light, and by the power of the telescope necessary to resolve star groupings into discrete stars. Then it was that he showed how — if we so judge — some of the nebulae must lie at distances so great that light cannot come to us from them in less than thousands — nay, in some cases, hundreds of thousands — of years. Many of the stellar nebulae, estimated in this way, must be con- sidered outlying galaxies, as grand perhaps as our own, or even grander. But Sir W. Herschel did not live to test, as he had proposed to test, this particu- lar method of gauging the celestial depths. He made numbers of observations preparatory to the work of testing it ; he obtained multitudes of results which, judged by the suggested criterion, possessed a most marvellous interest ; but he was not able to test the criterion, as he had before tested his first gauging method. That first method, it will be remembered, was based on the assumption that the stars which form our THE NEW STAR. II Stellar system are strewn with general uniformity alike as to size and number throughout all parts of the system ; and it was by this criterion that he pro- posed to measure the extension of the galaxy in different directions. Had he died after collecting his gauges, without using them to test his criterion, it might have remained an accepted truth that he had been content with it. Fortunately, besides col- lecting those valuable results, he examined and in- terpreted them, announcing in so many words : " I am now convinced, after my long and careful exami- nation of it, that the stars which form the Milky Way are very differently arranged from those in our neighbourhood." Had Sir W. Herschel lived to test his second method* of gauging the star-cloudlets, he would have seen that the criterion on which it depends is unsound, the results to which it leads being incon- gruous, as will presently be seen. But unquestionably the results which Sir W. Herschel published in 1817 and 1818 justify the belief that, as interpreted by his criterion, large numbers of the nebulse must be regarded as external galaxies. This grand conception fascinated the minds of astronomers, and of some who, like Humboldt, though not themselves astronomers, yet had under- * So carelessly have Herschel's papers been read — or skipped through — by those who most confidently quote them, that the very existence of two methods of gauging in his descriptions of his plans has not been noticed. When the point is men- tioned, answer comes readily that the two methods are prac- tically the same. How far this is the case will be seen when I note that in the first the same telescope was used throughout, the gauging test depending on the number of stars seen in different directions ; while in the second telescopes of many degrees of power were used, the gauging test depending on the resolution of one and the same star grouping. One method was employed because the other had failed. Yet in some books the result obtained by both methods are hopelessly mixed together. 12 OTHER SUNS THAN OURS. Stood and appreciated the work of the great observer. The idea of "island universes " strewn throughout the ocean of space impressed the world by its over- whelming suggestions of vastness alike in space and time and energy. It was Mr. Herbert Spencer who first showed the incongruity of the results which Sir W. Herschel had collected in those papers. On the one hand, we learned that our galaxy is in parts absolutely un- fathomable, insomuch that even with Herschel's most powerful telescopes there still remained regions of irresolvable milky light, though with each increase of power more and more stars had been revealed. On the other hand, we were told of outlying galaxies, similar to our own in structure, but lying at distances many times exceeding its whole span, and these galaxies were found in many cases to be resolvable into stars. According to theveryprinciple on which the second method of gauging distances was based, these galaxies being resolvable are much nearer than the outer parts of our own galaxy. So that if the prin- ciple be accepted these resolvable star-clouds are not external : if they are external, the principle of the second method of star-gauging must be rejected, and all the results based on it must be re-examined. Another objection, almost as fatal, was noted by Mr. Spencer in the same remarkable essay (on the " Nebular Hypothesis "). Sir W. Herschel noticed that when in surveying the star-depths he came to regions where stars were few, nebulae were almost sure to be found. On such occasions he would call to his assistant (his elder sister, Miss Caroline Herschel) " Prepare to write, nebulse are about to ap- pear." Strangely enough this peculiarity has actually been adduced as affording evidence that the nebulae are outside our system, as though insects in a tree who should notice that as they pass from branches to twigs, and from the larger twigs to smaller ones. THE NEW STAR. 1 3 they came upon leaves, should regard this as afford- ing proof that the leaves do not belong to the tree. Of course, as Mr. Spencer points out, this peculiar relation proves, if it be shown to be a real and not an accidental relation, that the nebulae belong to the same great system as the various orders of stars. In the years 1867-70, unaware of Mr. Spencer's prior observations, I noted the same objections and several others. I made equal-surface charts, indicat- ing the distribution ot stars and nebulae. These showed in the clearest possible way that not only do the nebulae appear in small regions clear of stars, such as Sir W. Herschel was dealing with, but that on the larger scale also the same peculiarity is pre- sented. In my maps the streams of the Milky Way appear like the branches of a tree, the nebulae occu- pying the region around like the twig-work and the leaves of a tree around the branches. Classifying the nebulae, the peculiarity of arrangement was still more significant, and still better corresponded with the illustration I have just used. The coarser clusters were found near the Milky Way, as the larger twigs lie near the branches of a tree ; the closer clusters were farther away, while the irresolvable nebulae were found in greatest profusion far from the rich galactic regions. But the irregular gaseous nebulae were all found on the Milky Way, so that they might be compared to ivy clinging close around the great branches of our illustrative tree. But there is one region of the heavens where this relation is exactly reversed. In the two Magellanic Clouds, or Nubeculae, which look as if they were great masses that had floated off from the Milky Way, there are all orders of stars, from those just beyond the reach of the naked eye, down to star-clouds of milky aspect, which no telescope can resolve. In this respect the Magellanic Clouds resemble those parts of the Milky Way which Sir W. Herschel found r4 OTHER SUNS THAN OURS. unfathomable. Buj: in both clouds are found all orders of star-cloudlets. There is even an example of the great irregular gaseous type — the celebrated Lover's Knot Nebula in the Sword Fish. In my maps of nebulae the Magellanic clouds are actually pictured by the great gathering of nebulae over their areas. In this respect the Nubeculae are entirely unlike the Milky Way. Dr. Whewell (in his " Plurality of Worlds ") was the first to show that this peculiarity requires us to re-examine the theory that the nebulae are external systems. Their gathering on the Magellanic Clouds is certainly not accidental or only apparent, their real positions being far beyond the stellar parts of the Nubeculae. These great clouds are certainly rounded in real shape, and therefore their remotest parts not farther than their nearest in greater degree (determined from their apparent size) than as ten exceeds nine. Hence, within these narrow limits of distance, we have all orders of stars, from the seventh to the twentieth magnitude and beyond, and all the various orders of star-cloudlets. These star-cloudlets then, at any rate, are not external galaxies. Is this peculiarity inconsistent with the other, so that it tells a different story t On the contrary, it tells precisely the same story. In the segregation of nebulae from star-streams we have what corre- sponds to the view of a tree from within, as by in- sects, seeing leaves plentiful at some distance from the branches, but not close to them. In the aggre- gation of nebulae within the Magellanic Clouds we have what corresponds to the view of a tree from without, as by men who see branches, twigs, and leaves as all parts of one system.* * That we are outside the Magellanic Clouds is clear from their appearance — but proved in another way by Sir John Herschel's observation that " the access to the Nubeculae is on all sides through a desert." THE NEW STAR. IS To consider no further evidence, it had in reality been demonstrated, as I pointed out in my lecture on " Star Drift, Star Clouds, and Star Mist," at the Royal Institution in May, 1870, that the nebulae of the stellar sort are not external galaxies. But reasoning such as had hitherto been employed has very little influence even on students of science. The sun's corona had long been proved by reasoning, based on evidence already obtained, to be a solar appendage, before the success of the photographers during the total eclipses of 1870 and 1871 convinced every one that it is so. And in like manner Mr. Spencer's arguments and my own, demonstrative though they were, convinced few, being " caviare to the general." The new discovery just made, however, or rather the event which has just taken place, can be mis- understood by none. A new star in the midst of an external galaxy would require to be many millions of times larger than our sun to be visible as the new star in Andromeda has been. The changes of lustre shown by the star would signify changes of energy, in the way of increase and diminution, the least of which would correspond to all the energy shown by our sun during hundreds of thousands of years. Even admitting for a moment that the new sun might be such an orb as this implies, yet the existence of such a sun within the great nebula in Andromeda would of itself show that that star-cloud cannot be a galaxy in the slightest degree resembling our own. The new star tells us, then, that the Andromeda Nebula, as its position had long since shown the more observant, lies within the limits of our galaxy. This strange mass of matter, not gaseous yet not stellar, a vast cloud perhaps of cosmical dust main- tained at intense heat in some way as yet unknown, surrounded by other cloud-like matter capable of l6 OTHER SUN'S THAN OURS. intercepting the red rays, much as dust in our own air intercepted and reflected the red rays of the sun when we had those marvellous coloured sunsets, tells us of one of the varieties of constitution and aggre- gation existing within our stellar system. We have every reason for inferring that the new star appeared in a part of the galaxy which is probably the nearest to our solar system. In this part, from Orion to Cepheus, the five nebulae visible to the naked eye are all found. Two are shown in the illustrative maps (figs. I and 2), and these objects are probably as near to us as, on the average, are the lucid stars {i.e., those visible to the eye), if not nearer. The picture of the Andromeda Nebula (fig. 3) is from the view made with the Harvard telescope by Trouvelot. The new star is not quite central, but it is in the very heart of the nuclear region of the nebula. It is no external galaxy but a part — a strange and highly interesting but in reality a very small part — of our own galaxy. So it will now be admitted are its fellow nebulae of all orders. Our stellar system, or galaxy, presents itself then to us in a new aspect. Like a mighty tree it spread broad arms, the stellar branches and streams and closely gathering aggregations which form the com- plicated wealth of the Milky Way. But the main stem is found in the four orders of isolated suns, the giant suns like.Sirius and his fellows, the suns like our own, the suns which show signs of darkening vaporous envelopes, and doubtless multitudes of dead suns. These are the chief though not the most numerous bodies in the galaxy. Even among these we find other varieties, in single, double, triple, and multiple suns, separate suns of all colours, sun groups presenting the most singular and beautiful combinations of colour. But as we pass to the borders of the Milky Way we find other varieties of structure. Here we have diffuse clusterings of stars, Fig. 3.— The Great Nebula tn Andromeda. To/aci THE NEW STAR. XJ farther off we find definite clusters of many orders, the least compact and most readily resolvable being on the whole nearer to the Milky Way than those in which the texture is finer or closer, while the irresolv- able nebulae tend to the regions remotest from the galaxy. Other forms of star-cloudlet there are also, which cannot be described as mere clusters. Then there are all the varied orders of real nebula — ring- shaped, spiral, planetary, and irregular. Doubtless, also, as our survey \i continued, fresh forms of structure will be recognized, till men are disposed to smile at the cloven flat disc of uniformly strewn suns which has so long done duty for our amazingly complex galaxy. We may still believe in external galaxies, though none may be within the ken of our telescopes, incompetent a§ yet to reveal the full extent and the full glory of our own star-system. But those external galaxies do not repeat the uniform and scarcely interesting galaxy we formerly judged ours to be, but our galaxy as it really is, infinitely complex in structure, immeasurable in extent, and to our conceptions full of infinite and everlasting vitality and energy. CHAPTER II. THE BIRTH OF WORLDS. The new star in Andromeda has been popularly regarded as probably a new world. This, whatever else it may be, it assuredly is not. In like manner the new star in the Northern Crown was popularly regarded (by persons unversed in science) as a world in flames. Stars are of course not worlds, whether they be new or temporary or simply variable. The idea gains ground steadily that all so-called new stars — even the glorious orbs seen in remote times, which outshone Sirius in splendour — were but variable stars, with a somewhat exceptional range of variation, and probably of very long period. If the star Mira or Wonderful, in the constellation Cetus, were so situated that when at its faintest it was visible as a third-magnitude star, it would out- shine all the stars in the heavens when at its ■maximum of splendour. So would Eta Argfts, and so also would the so-called New Star in the Northern Crown. Indeed, if we regard the nebula in An- dromeda as lying farther away than the faintest star visible to the naked eye, then, were we brought so much nearer that its distance was only that of a first-magnitude star, the nova stella (probably but a Stella mutabilii) which shone out recently in its midst would have been resplendently visible instead of needing a telescope for its detection. Neither this star, nor any other new variable, or temporary star ever observed, can be said to have i8 THE BIRTH OF WORLDS. I9 thrown the least light on the birth of worlds. Cer- tainly, if the nebular hypothesis of Laplace repre- sents the real way in which solar systems are formed, no new star has thrown light upon that process, or possibly can ; for the process imagined by Laplace involved no catastrophes. It was a steadily acting process, rather leaving nebulous rings behind than throwing them off as commonly supposed ; the rings separated into parts as they shrank longitudinally by a gentle movement, and the various fragments coalesced rather than collided, for they were all travelling the same way round ; in fine, Laplace imagined no fierce conflict of matter with matter such as the sudden outburst of splendour in what we call a new star necessarily implies. It might be well, however, if the interest excited by the new star, though it may throw no new light on Laplace's hypothesis, should direct some degree of attention to the very remarkable defects which any astronomer who knows aught of physics, or any physicist who knows much of astronomy, cannot fail to recognize in that remarkable speculation Attracted by the effective way in which some features of our solar system for which the theory of gravi- tation cannot account, appear to be explained by Laplace's hypothesis, many astronomers overlook the startling difficulty which Laplace overleapt at the outset. On the other hand, many physicists are unaware that the hypothesis started from what, with the knowledge of physics obtained since Laplace's time, is seen at once to be an absolute impossibility ; they know only that a number of astronomical facts appear to require some such theory ; of the details which are also required (but which a physicist at once sees to be quite impossible) they know little. Let us consider how the theory of Laplace was suggested and what the theory required, premising that if the basis of the theory shall appear to be C 2 20 OTHER SUNS THAN OURS. more than unstable that involves no discredit to Laplace, seeing that in his days certain physical laws which are now among the axioms of science were not even suspected. We may take, as an example of what Laplace could and could not do, that masterpiece of mathematical analysis, his inquiry into the stability of Saturn's ring-system. Here the mathematical work was almost perfect, and the conclusion, that the rings must be narrow and eccentrically weighted, was demonstrably right, on the assumed premisses ; but these premisses were erroneous. A knowledge of physical laws such as Laplace could not have, but such as many boys in our times have acquired, would have shown Laplace that the rings of Saturn could not be what he assumed them (quite unquestioningly) to be, at the very outset of his inquiry. Solid rings on the scale of the Saturnian system could no more remain unbroken under the forces to which they are subjected than a model of the Menai Bridge, perfect in all other respects, but on such a scale as to span lOO miles, could bear its own weight. In this case, where not a theory, but a magnificent calculation of his, was in question, science has not hesitated to set Laplace's conclusions aside, because of the falsity of his assumptions, adopting, instead, the results which Clerk Maxwell, Pierce, the Bonds, and others have established — viz., that the Saturnian rings consist of myriads of tiny satellites, like sands on the sea-shore for multitude. But, strangely enough, in the case of his far-famed (chiefly because so imposing) hypo- thesis of the birth of worlds, which starts from a similar, or rather from a much more monstrous, mistake (very natural, though, in Laplace's time), science has scarcely even questioned his results, far less examined his initial assumptions. The facts which the nebular hypothesis of Laplace was intended to explain are simply these : — The planets travel the same way round, and in nearly the THE BIRTH OF WORLDS. 21 same plane (all but the zone of minor planets, whose entire mass is less than the tenth part of our earth's). The central sun turns the same way on his axis, so do all the planets whose rotation has been observed ; all the moons travel round their ruling planets the same way — except the moons of Uranus (known, it must be remembered, to Laplace) and the moon of Neptune ; and tiiese bodies, travelling as they do at the very outskirts of our system, may be regarded as having, perhaps, been exposed to disturbing influ- ences affecting, in their case, the action of the laws, whatever they were, which gave these features of uniformity to our solar system. Laplace suggested, as a hypothesis which seemed to him to result {une hypothise qui me taralt rhulter) from these features, that the whole mass of matter out of which the solar system was formed was once an immense disc, ex- tending beyond the path of the remotest planet now known, and rotating as one gigantic whole. Grant- ing only this assumption, and starting from it, all the features of the solar system mentioned above would follow. The ring would gradually shrink as its heat was radiated into space, until the outer parts, retain- ing their original velocity, could no longer cohere, but would be left outside in the form of a gigantic ring. This ring, as it further shrank (along its length now), would dissolve into fragments, and these would eventually coalesce into a single planet, the outer- most. Then another would be formed in the same way, and another, and yet another, until at last there would be left, in the middle, the great mass that was afterwards to govern that family of worlds. Each planet, at its beginning, being like the original gase- ous disc, would go through a similar process of con- traction, and form as many bodies subsidiary to itself as its quantity of matter and the conditions under whidi it had itself been formed would allow. The process would fail in some cases, and so several small 22 OTHER SUNS THAN OURS. planets would be formed instead of one large one, as we see in the case of the asteroids ; or, as in the case of Saturn's system, a ring or set of rings (rings of small satellites, as we now know) would form instead of a single large satellite. Laplace's theory, if we grant its initial assumption, accounts fairly for all the features of the solar system, except the singular distribution of the planets into families : — the giant planets outside, as if guarding the rest ; the terrestrial planets near the sun, as if under his protecting wing ; and the asteroids or minor planets in the mid space between these families, as if keeping them apart. But unfortunately the initial assumption, on which the whole theory de- pends, is as utterly inadmissible as the theory that Saturn's rings might conceivably be solid. It is almost inconceivable how amazingly impossible that initial assumption is. Few probably know that a solid disc of steel, extending only to the earth's orbit, could not move as a single mass. If the cen- tral part of such a disc — say a region as large as the sun's globe — were set rotating as by some mighty hand, the outer parts would not feel the impulse until more than ten months had elapsed. But imagine a disc extending to the orbit of the planet Neptune> thirty times farther from the centre than is the earth's path. Imagine, further, such a disc-shaped region of space, not occupied by a mighty mass of the stoutest steel, but by a vaporous mass many thousands of times more tenuous than the air we breathe. It is such a disc that we have to imagine, according to Laplace's theory, rotating as a single mass. No argument is really needed to show that this is abso- lutely impossible. But it is a truly remarkable circumstance that, while a mathematician like Clerk Maxwell did not hesitate to point out (with perfect justice, be it remarked) that the solid flat rings which Laplace recognized in the Saturnian system, because THE BIRTH OF WORLDS. 23 they seemed to be plainly visible there, would be absolutely plastic under the forces to which they were exposed, astronomers and physicists have been apparently afraid to acknowledge that a vaporous disc such as he only imagined, a disc rarer than the rarest known gas, so vast that the whole Saturnian system would be but as a speck by comparison, and moved by far mightier forces than act on that sys- tem, could have no coherence whatever, and could not possibly even begin to behave as Laplace's theory required. If the mere mathematician had been thus weak, we might not have wondered, for mathematicians often rejoice over problems depend- ing on impossible conditions — perfect rigidity, abso- lute uniformity, entire absence of friction, and so forth. But physicists and astronomers have usually required conditions more in accordance with the actual workings of nature. CHAPTER III. WILLIAM HERSCHEL'S STAR SURVEYS. " Coelorum perrupit Claustra." — HerscheVs Epitaph Among the researches that I should wish to live to see undertaken by astronomers, and especially by astronomers capable of applying photographic methods to the work, I regard with particular interest the survey of the stellar depths, in accordance with the original ideas of Sir William Herschel, but on principles such as he by no means supposed to be correct when he began his labours. It has been unfortunate for the work of research in this direc- tion that Herschel's ideas and results during forty years of observation have been dealt with, by astronomers who came after him, as though they had been presented in a single treatise and indicated his views at some one given time. In a sense, there is something singularly appropriate to the grand subject with which he dealt in the particular quality of the picture that we have received from his hands. The starlit heavens present a similar diversity in regard to time. We find it difficult, nay, impossible, to conceive that the stars as we see them are not as they actually are, nor even as they were at any given time. We do not see any star in its true place, even after correction has been made for. such effects as are produced by atmospheric refraction and aberra- tion of light For each star is rushing swiftly through space, changing its apparent position in the 24 WILLIAM HERSCHELS STAR SURVEYS. 25 celestial sphere, and although, owing to the enormous distance of each star, the apparent movement is not perceptible by ordinary eye-sigfat in less than hun- dreds of years, yet as ligh't takes many years in reaching us from any star, it remains strictly true that the position apparently occupied by a star is not its real position, but one that it occupied long ago. Again, we do not see any star with its real light at the moment, but with that (by no means necessarily the same, even in amount) which it emitted many years ago. Even this is difficult to conceive; but this is little. Each star tells us of its history at a particular time, corresponding to its distance. Yonder bright star shows us its position and lustre a score of years since ; the less brilliant orb apparently close by, lies so much farther off that we must assign the news it brings us to at least a century ago. But many even of the brighter stars lie at much greater distances ; while, when we pass to the fainter stars, we must often have to consider light-journeys of many hundreds or even many thousands of years. If we regard the telescopic view of the heavens as the real view presented to the eye, — at least, to the mind's eye of science, — we must recognize, in the case of the faintest stars seen by the most powerful telescopes, such vast distances that light cannot have come to us from those stars in less than hundreds of thousands, perhaps millions of years. So that the scientific view of the universe of stars has as wide a range in time as in space. We have no picture of the galaxy as it actually is, or even as it was, but of different parts inextricably intermingled, and at different, and very widely different, periods of time. But science enables us to correct the mistaken idea that in the stellar heavens we see the universe of stars as it is at this very time. Though the mind may never be enabled to conceive the reality, and is, indeed, hopelessly unable even to approach the con- 26 OTHER SUNS THAN OURS. ception, yet the reason has been convinced long since that the stellar heavens tell the amazing story of vast realms of space and enormous durations of time, which modern astronomy has in part been able to read. It is not very wonderful, but it is interesting and significant, that the labours of the man that has done most to bring the great problem of the star-depths before us should have been misinterpreted somewhat as we are so apt to misinterpret the heavens them- selves. Writers even so able as Humboldt and Arago take statements from this and that part of Sir William Herschel's long series of papers, and set them side by side in the same page, or even in the same paragraph ; nay, I have seen such statements wrought into a single sentence, when in reality they belong to entirely different parts of Herschel's pro- cess of inquiry, or even present entirely distinct views on the particular matter to which they relate. Although my chief work has long been to try to put myself in the position of those who are apt to make mistakes, in order that I may be the more successful in correcting such errors, I still marvel how so gross a mistake can have been made. Sir William Herschel suggested, in the course of his career as an observer of the stars, two entirely distinct methods of gauging the star-depths. They were so different in character, that — to take but one point of difference — one depended on the use of one and the same telescope throughout, while the other required that a series of telescopes of gradually increasing power should be employed. Yet, not only have superficial readers overlooked the characteristic difference between the two methods, and the reason why one method gave place to the other, but even those that have professedly under- taken the work of analyzing and abstracting the labours of the great astronomer of Slough, have fallen into the same preposterous mistake. I know WILLIAM HERSCHEL'S STAR SURVEYS. VJ of only one, Wilhelm Struve of Pulkova, who has clearly recognized and insisted upon the difference between the two systems of space-gauging that were employed by Sir William Herschel at the beginning and toward the close of his marvellous series of observations. Even Struve failed to recognize clearly that Herschel never did more than sketch in outline the results that would have followed from his second method of gauging, interpreted in a way that seemed to him likely to be sound and just. Herschel was too old to do more ; and, apart from this, it may be said that he left those who came after him not only to apply the method fully, but even to interpret satisfactorily the few results that he had himself been able to collect. Every one knows the nature of the system of star- gauging that Herschel at first adopted ; in fact, it is the only one about which the great majority of students of astronomy know anything. It was the method suggested originally by Wright of Durham; Supposing all the stars visible in the telescope to belong to a certain system of stars tolerably uniform in size and distribution throughout (our sun being one of them), it is easily seen that if the telescope we use brings into view all parts, even the remotest, of this star-system, we can determine the shape of the system with considerable accuracy. For, in whatever direction we turn the telescope, we shall see a number of stars, greater or less according as the boundary of the stellar system in that direction is farther or nearer. Wright of Durham applied this method of gauging, with a telescope of moderate power, with results closely resembling those that are presented to this day as among the chief triumphs of Sir William Herschel's entire series of labours;. Wright found so many stars in the direction of the Milky Way, compared with the numbers seen in those parts of the sky that are free from milky light, 28 OTHER SUNS THAN OURS. that he was forced to assign a much greater exten- sion to the stellar system in the direction of the Milky Way than elsewhere. Forced, at least, when we consider the assumption on which his inquiry had been based ; for of course there were several other available explanations of the observed facts. Thus Wright was led to enunciate the theory, com- monly attributed to Sir William Herschel, that the stellar system has the shape of a gigantic flat disc of stars, tolerably uniform in distribution. The Milky Way being divided into two streams along a part of its course as known to Wright, it was necessary to assume that the disc was cloven throughout half of its extent. Sir William Herschel, making a more careful sur- vey on the same plan, but with a much more power- ful telescope, found that while in a sense this cloven flat disc theory was supported by the results he obtained, it was yet necessary to assign a much more complex figure to the stellar system, so long as the results of his gauges were interpreted in accordance with the assumptions suggested by Wright It became clear that on these assumptions the bounding surfaces of the flat star-system were by no means smooth. Instead of a section of the stellar system through its centre (near our sun) and at right angles to its median plane being bounded by straight lines, the outline must be of the most irregular form. Herschel drew one of these sections, which presented a shape somewhat like that of a long, dentate leaf. He appears not to have been at all struck by the peculiarities of outline thus presented, when he was considering only a section of the stellar system. It is obvious that a system of stars forming a sort of island universe might be expected to present many irregularities of shape, and a section athwart the naiddle of such a system might as probably be shaped like a toothed leaf as in any other way. WILLIAM HERSCHEL'S STAR SURVEYS. 29 But as the work of survey went on, Herschel began to find that not only particular cross-sections, but the system itself, presented peculiarities of form, and that these were related in too special a way to the position of the observer on the earth to be easily explicable as really belonging to the system of stars. Consider, for instance, such a case as the following : Over a certain region of the heavens, nearly circular, Herschel found that his star-gaugings invariably gave high numbers, while over the region all around this nearly circular space they as systemat- ically gave very low numbers. Thus, if we suppose A, B, D, E to represent such a circular region, having Fig. 4. its centre at C, Herschel found that within the boundary A, B, D, E he always had fields of view rich in stars ; while so soon as he directed the tele- scope to points outside of A, B, D, E, he found not more than perhaps four or five stars, instead of hundreds, in each field of view. The meaning of this result — if the assumptions adopted by Wright and Herschel are accepted — is obvious. Herschel himself never hesitated in recognizing this meaning ; yet those who quote Herschel constantly, and regard with intense disfavour the idea that he could, under any circumstances, have made a mistake about the 30 OTHER SUNS THAN OURS. stellar universe, overlook the direct result of his observations, the result pointed out by himself and frankly accepted. If within a small circular or roughly rounded space, such as A, B, D, E, many stars can be counted in every field of view, while over the whole space L, M, N, K, outside of A, B, D, E, few stars are seen, and if a great number of stars seen in any direction indicate a corresponding!}- great extension of the stellar system in that direction, then, of course, it follows inevitably that the stellar system extends toward the region A, B, D, E very much farther than toward any of the region around A, B, D, E. If the stars over the space A, B, D, E were uniformly distributed, the conclusion would be that a cylin- drical projection or rod-shaped extension of the stellar system existed in the direction toward C, the centre of this rounded, rich region of stars. If, on the other hand, as Herschel found to be almost invariably the case, the stars, though rich over the whole region A, B, D, E, were much more closely aggregated near the centre, C, than toward the edge, the conclusion would be that there was a conical projection of enormous length compared with its breadth, having its axis directed toward C. In neither case could the conclusion be regarded as reasonably likely, scarcely even within the bounds of probability. It would be strange enough to imagine a star-system of vast extent, with long cylindrical or conical projections extending from portions of the central group, the extensions being many times longer than the diameter of the parts of the central mass (the cloven flat disc of stars) from which they sprang. Nay, this would not only be strange, but altogether inadmissible when dynamical laws are taken into account. But if we overlook the strangeness and the unscientific nature of such a supposition, we find another and overwhelming WILLIAM HERSCHEL'S STAR SURVEYS. 31 difficulty in the peculiarity that every one of these strangely projecting cylinders and cones of stars must be conceived as having its axis directed exactly toward the solar system, from a member of which we make our observations. Our sun is, by the very assumption on which the system of numerical star- gauging depends, but one among millions of suns forming a system of stars. There is no reason whatever for supposing that he lies at the centre of the system, or, indeed, that the system is of such form as to have a " centre of figure," which, of course, can only exist in the case of a symmetrical system. On the contrary, there are abundant reasons in the complex form and various degrees of brightness of the Milky Way, and in the general superiority of lustre found within its southern portions, for believing that the system of stars (if, indeed, the Milky Way represents its richer parts) is exceedingly complex in shape, and the sun eccentrically placed within its limits. Yet, as seen from this casual star — for in looking from the earth we get, to all intents and purposes, the same view of the stellar depths as if we looked from the sun — all the strange projecting spikes of stars (I can think of no more suitable name) are foreshortened into the appearance of round star-clusters ! This is absolutely incredible. There can be no doubt or question as to the sig- nificance of the observed facts, if the assumption on which the star-counting method depended is accepted, and it is scarcely possible to entertain any doubt or question as to the absolute inadmissibility of the result thus obtained. If the greater the number of stars seen in any field of view, the greater is the extension of the star-system in the direction of those stars, there must be enormous spike-shaped pro- jections of stars wherever clustering aggregations are seen, along the Milky Way or elsewhere ; while the existence of such projections, always directed 32 OTHER SUNS THAN OURS. exactly toward the sun, cannot be admitted as possible by any reasoning mind. Sir William Herschel, at any rate, felt no doubt on the subject. He saw at once, that since the principle he had assumed in the beginning of his star-gauging by the counting method led to a result that was manifestly preposterous, the principle that had seemed so reasonable must be rejected as un- sound. Repeatedly we find him saying that a long continued examination of the star-system has con- vinced him that the idea of uniformity of distribu- tion, which he had imagined at the beginning, must be given up as inadmissible. He remarks that he has satisfied himself that the stars in the Milky Way are distributed very differently from those in our neighbourhood. He understood the real meaning of the clustering aggregations of the stars along the Milky Way, regarding these as manifestly real clusters of stars, not stellar projections. It would indeed matter little if Herschel had failed to recognize the meaning of what he had himself observed. Had he so failed, we should have found but another instance among hundreds known to us of the inaptitude of even the keenest observers to analyze their observations, and educe the full meaning of what they have discovered. Herschel differed from the rank and file of mere observers — the working army of science — in the power he possessed in this respect, until approach- ing the end of his wonderful observing career. But had he in this case failed to reason right — as in later years we find he actually failed — this should in no sense influence our judgment respecting facts that are as clearly before us as they were before him. We know that the assumption he first adopted would compel us to assign to the star-system a shape that is antecedently unlikely even as a shape, and is rendered utterly inconceivable when we take into WILLIAM HERSCHEL'S STAR SURVEYS. 33 account the peculiar relation of all its most marked features to the sun. If we saw a number of grains scattered over a surface at random, and found that as they fell they arranged themselves in the form of a star, all the radiations of the star-form being directed exactly toward a certain mark on the sur- face, we should be absolutely certain that there were peculiarities in the surface, differences of level, or the like, which brought about this result. It could not possibly be accidental. We ought to feel as certain that there cannot be multitudinous radiating streams of stars, all extending straight from our sun, unless there is some special peculiarity in our sun to cause this singular conformation of the star-system. And since we know certainly that no such peculiarity exists, we cannot but reject decisively the belief that the star-system is so shaped. It could make no difference whatever in our conclusion that Sir William Herschel had failed to notice the inference directly deducible from his observations. But, as a matter of fact, the elder Herschel accepted the rich clustering regions along the Milky Way as in reality what they appeared to be — that is, as clusters, not as projecting streams of uniformly strewn stars. Of course, the principle that he has assumed as the basis of this system of star-gauging — the prin- ciple of generally uniform distribution — had to be abandoned in at least these special cases. Prob- ably Herschel was not prepared to admit that it must be given up altogether. This seems much clearer in our time, with our vastly increased know- ledge about the stars, than it could have been to Herschel, keen though his insight into such matters unquestionably was. But Herschel went on at this time with a series of sidereal observations of the widest scope and the most diverse character. He had practically the whole field of stellar and nebular research ; the D 34 OTHER SUNS THAN OURS. universe was all before him where to choose, a noble but truly a bewildering scene. So far as obser- vational work was concerned, he could hardly go wrong, let him undertake what portion of the survey he might. Again and again he sent to the Royal Society the results of fresh series of observations — now a thousand or so of new nebulae discovered by him in his " sweeps " of the mighty dome of the heavens ; anon the survey of regions containing hundreds of thousands of stars ; then an inquiry into the distribution of nebulae and stars : and all this work went on in company with the observation of sun, moon, planets, and comets, the construction of new telescopes by hundreds, the study of many complex physical problems, and other scientific inquiries of minor importance. As these labours went on, and clearer ideas of the constitution of the heavens presented themselves, Herschel must have begun to see that the system of gauging the galaxy by counting stars was utterly inadequate. With all the various orders of star- clusters and nebulous masses, how could he longer imagine that mere numerical wealth of stars, or of points of light looking like stars, indicated enormous extension in the direction of the line of sight toward such regions ? Distance, indeed, he felt to be indicated by the close aggregation of multitudinous points of light. But the vast distance that he recognized in some of these clusters of stars was something entirely different from the long array of stars in particular directions that he had originally assumed as the explanation of great wealth of stars in such directions. His original idea of the structure of the stellar universe had not included the con- ception of star-clusters, either of the larger sort, such as he had found in parts of the Milky Way, or of the smaller kind, rounded, elliptical, irregular, ring-shaped, or of other forms of small clusters, which WILLIAM HERSCHEL'S STAR SURVEYS. 35 sometimes he was disposed to regard as external, stellar universes, at others as fragmentary portions of our own galaxy. It was a natural outcome of such observations as these, and of the doubts they inevitably cast on Herschel's original method of star-gauging (or rather of the conviction forced upon him that the principle of that method was untrustworthy), that Herschel should be led to devise another method. I wish specially to show that the method he now adopted was entirely different from the other, insomuch that it is among the marvels of misapprehension which the study of science brings before us that this method should be confounded with the earlier system. But I wish also to show how naturally the new method of star-gauging arose out of the observations on which Sir William Herschel had been engaged since his earlier star-gauging had shown him that the universe of stars is not constituted as at first he supposed it to be. Let us take the latter point first. Among the nebulae Herschel had found all orders of what he called " resolvability." Some of them are clusters so coarse in texture that it was not easy to draw a line of distinction between them and the more clustering portions of the galaxy itself I may notice in passing a feature that was not known to him, viz., that the nebulae of this coarsely clustering type are more numerous upon and in the neighbour- hood of the Milky Way than over the rest of the heavens. Others, again, are compact clusters, still easily resolved into stars with a telescope of moderate power. Then there are others so difficult to be resolved into stars, that until powerful telescopes were applied they presented the appearance of round or elliptical cloud-like spots. Yet others are still finer in their starry texture, so that only a few of the most powerful telescopes in the world will re.solve D 2 $6 OTHER SUNS THAN OURS. them into discrete points of light. And lastly, so far as the nebula of regular shape are concerned, there are some that have not yet been resolved into stars by any telescope. It is noteworthy that, arranging the nebulae into classes in the order of their resolvability, those most easily separated into stars show the most marked tendency to aggregation along the Milky Way, and are irregular in shape. Those that come next in order are nearly circular, and though still showing a certain increase of wealth toward the Milky Way, are found in tolerable frequency elsewhere over the star-sphere. The nebute that are resolvable with difificulty, on the other hand, are elliptical, and are absent altogether from the Milky Way. These points are manifestly associated with the great problem of the constitution of our galaxy, though not directly related to Sir William Herschel's observations. In fact, though he noticed the remarkable circumstance that the nebulae cluster near the northern pole of the Milky Way (that is, near the point farthest on the northern heavens from the central line of the Milky Way), iie did not recognize the manner in which this peculiarity is associated with the character of the nebulae, and he supposed that the nebulae are rich along a ring-shaped region akin to the Milky Way, but at right angles to it, and formed of star-clouds instead of stars. Recognizing these diversities in the structure of nebular^, Herschel was naturally led to regard them as due to differences of distance. He supposed the coarser clusters to be the nearer, and the finer in stellar texture to be the more remote. All nebulae might fairly be regarded, at that stage of the inquiry, as farther away than the stars forming our own sidereal system, even to the farthermost parts of the galaxy. Herschel does indeed speak of the possi- bility that toward the side of our flat sidereal system, WILLIAM HERSCHEL'S STAR SURVEYS. 37 as he viewed it, there might be room for the nearer approach of the parts of a former great single nebula, as though the nebulae seen clustering in great numbers over the wings, shoulders, and head of Virgo might be but the parts of a former nebula of gigantic proportions. But this notion seems not to have been more than a passing idea with him, or to have much influenced the development of his views. The idea gradually gained force, on the contrary, that in the greater or less telescopic power needed to resolve a nebula, or a group of stars, we may find evidence of the greater or less distance of the object so scrutinized. So soon as this idea had taken firm root in his mind, which was not till toward the end of his observing career, he proceeded to put to use this means (as he supposed) of determining distance. I refrain from saying that he put it to the test, for I have no evidence that he consciously did so. He seems to have taken it for granted that the visibility of a star as a separate point of light, by a telescope of given power, was in itself a test of distance. He stated the principle, and showed how it might be applied to stars, star-groups, star-cluster.s, and nebula; of various orders ; then he proceeded to employ it as a means, first, of measuring the scale on which the stellar system is constructed, then of determining its shape, and lastly, of ascertaining the distances ot the nebulae. And now to show how entirely distinct was this method of gauging the star-depths from that which Sir William Herschel had before employed. We may call the first method star-gauging by enumera- tion ; the second, star-gauging by resolution. In the first method, the same telescope (a powerful one) was to be applied to different parts of the star-depths, the number of stars counted, and, as the number was greater or less, the limits of the stellar system in the given direction was assumed to be farther away 38 OTHER SUNS THAN OURS. or nearer. What was taken for granted in this method was, first, that the stellar system is formed of stars generally uniform in distribution throughout the system ; secondly, that the telescope employed was powerful enough (it was eighteen inches in dia- meter) to reach to the limits of the system ; thirdly, that there are no vacant spaces in the system. In the second method, different telescopes, ranging in power from the weakest in use to the most power- ful he could make, were directed to each region ex- amined, until the whole region had, if possible, been resolved into stars well defined on a black background, without any trace of milky nebulosity. What was assumed in this method was, first, that the sidereal system is formed of stars not differing greatly from one another in size ; secondly, that in the various clustering regions throughout the sidereal system the average distances between stars are tolerably uni- form, or, in other words, that what may be called the stellar texture of each part of the system is the same throughout, though there may be vacant spaces in some parts, and clustering aggregations of various forms in others ; thirdly, that any part of the system that the most powerful telescope he employed failed to resolve, lay at a distance beyond the gauging or fathoming range of that telescope. To bring the two methods more clearly into contrast, note that In Herschel's first method In Herschel's second method of gauging, it was essential of gauging, it was essential that that one and the same tele- a series of telescopes differing scope should be used through- in power should be employed, out the work. The comparisons made re- The comparisons made re- lated to different fields of view, lated to the same field of view, seen with the same light- seen With dififerentlight-gather- gathering powers. ing powers. The inference deduced re- The inference deduced re- lated to the extension along lated to the distance of objects the line of sight of the objects seen with the different tele- counted with the one telescope scopes employed, employed. WILLIAM HERSCHEL'S STAR SURVEYS. 39 Had Herschel been a younger man when he thought of the second method of gauging the star-depths, it is probable he would have felt from the beginning that the method was one to be tested before it could be trusted. He would have been prepared to find that while, if his assumptions were sound, his results would have such and such a meaning, it was at least possible that his results might show that his assump- tions were altogether inadmissible, and therefore that his new method of star-gauging was altogether un- sound. But Herschel was nearly seventy-nine years old when he began to employ his second system of star-gauging, and though he still possessed much of his skill as an observer, he had lost much of that versatility of mind which had enabled him not only to observe skilfully, but so to analyze his results as to see whether they were consistent with the assump- tions by which they were to be interpreted. Can we wonder if at that advanced age Herschel was content to work resolutely at the task on which he had entered, without considering very closely or thoughtfully the question whether the principle by which he proposed to interpret his results was sound or otherwise ? It had seemed to him so reasonable as to appear almost unquestionable ; we do not find a line or a word tending to show that he ever ques- tioned it. The principles on which the first method of star-gauging had been based had seemed to him equally unquestionable at first ; but he had found them to be unsound by noting that his observations interpreted by means of them led to absurdities. The observations made in accordance with the second method of gauging led in like manner, if interpreted by means of the principles on which that system was based, to absurdities. But this, his attention being directed too exclusively to the results themselves, he failed to recognize. Herschel began this new work of star-gauging by 40 OTHER SUNS THAN OURS. examining individual stars. It is clear that the principle of the method is applicable to a star as readily as to a star-cluster. If we can determine the average distance of those stars that we can just see with the naked eye on a dark and clear night, and stars generally throughout the stellar system have the same mean size (by which I mean that the average for a thousand stars in any one part of the system is the same as for a thousand stars in any other part of the system), then, of course, a telescope increasing the light-gathering power of the eye four- fold will just show a star twice as far away ; one increasing that power ninefold will just show a star three times as far away, and so forth. It was by observations made in this way that Herschel was led to the belief that among the stars shown by his most powerful telescopes are some that are thousands of years' light-journey from the earth. Singularly enough, the very evidence that shows in this case that the principle of the new method of star-gauging failed, has shown that the same result can be inferred that Herschel based on that principle. We know now, for example, that many of the brightest stars — as Sirius, Capella, Vega, Arcturus, and Aldebaran — are much farther away than some — as 6i Cygni — that are barely visible to the naked eye on the darkest and clearest night, instead of these being (as they should, if the principle of the new method were sound) fully a hundred times farther off. We can- not, then, any longer assume, as Herschel did, that the faintest stars seen with his largest telescopes are thousands of times farther away than those forming our constellations. They may be relatively near, and look small because they really are much smaller than their fellows. But while, on the one hand, we cannot now suppose faint stars necessarily far away, we are precluded, on the other hand, from inferring that bright stars are necessarily near. Since it is WILLIAM HERSCHEL'S STAR SURVEYS. 4I certain that many of the brightest among the stars visible to the naked eye are really farther away than many of those that are barely discerned, we may infer, with considerable confidence, that the same holds in the case of the field of view of the mightiest telescope yet made. Now, the faintest stars seen in such a field are those that would be the brightest in fields of view obtained by penetrating still more deeply into space. Among them, therefore, must be some farther away than those yet fainter stars ; among them, in fact, are probably stars like Sirius, Canopus, and Alpha Centauri, which owe their brightness to real vastness, and lie at depths remoter than the daring conception even of the elder Herschel had sugge'sted. But it is when we turn to the study of star-clusters that we recognize at once howthoroughly the principle of the new method of star-gauging was disproved, and how important, nevertheless, are the results that Herschel's observations on the new plan established. If he had found that each cluster, whether in the Milky Way or of the nature of a star-cloud, had been resolved by the application of a certain telescopic power, or of powers ranging between tolerably close limits, he might logically have been content to believe that his principle was sound. An easily resolved cluster would be set relatively near, and one resolved with difficulty would be set far away. But, as a matter of fact, he met with a very different result in many cases ; and a single case of the kind would have sufficed to dispose of the principle he had adopted. He found clustering regions (rounded in form) that were partly resolved by even his weakest telescopes, and more and more resolved on each increase of telescopic power, until he brought into action his very largest telescope ; but even with this instrument, milky nebulosity still remained. This peculiarity would be limited to a certain rounded 42 OTHER SUNS THAN OURS. space, in some cases not so large as the disc of the full moon.* Nichol says of these regions, "What wonder if even Herschel shrank back appalled in the presence of these unfathomable abysms .' " Herschel himself spoke less turgidly. He simply says, "When I have been unable to resolve the Milky Way with my most powerful telescopes, it has been because the Milky Way is unfathomable."t Now, this observation, interpreted by the principle of the second method of star-gauging, leads to pre- cisely the same absurdity to which Herschel had been led by his first method — and still more definitely, though not quite so obviously. All round one of these regions that he found unfathomable, the star- depths were easily fathomed, and therefore in those directions the stellar system had no great extension. But in the direction of these unfathomable regions the star-system had an enormous extension, if the principle of the new method could be trusted. The case is precisely the same as though a surveyor of the depths of ocean found that all over a large area of the sea bottom, save one spot, a few yards perhaps in length and breadth, he reached bottom with a hundred fathoms or so, while at that spot he could * Herschel himself does not dwell on this particular point, though it could not possibly have escaped his attention ; but any telescopist can ascertain for himself that all round the "unfathomable" regions noted by Herschel are regions that even moderate telescopic power will completely resolve. + Struve was led into a singular mistake by this sentence (which I quote from memory, but. correctly in essentials). He wrote it out probably in German, " Wenn Ich," etc. ; or if not, he simply understood it as if the English word_" when " were equivalent to the German " wenn " ; for in his " fitudes d'Astro- nomie Stellaire " he writes the sentence with the word " Si " for "when," making the statement, which Herschel applied to those parts only of the Milky Way that he could not fathom, relate apparently to the whole of the Milky Way, and suggest- ing consequently an infinitely extending flat, galactic disc fof Herschel's finite one. WILLIAM HERSCHEL'S STAR SURVEYS. 43 not reach bottom with a line of two or three thousand fathoms ; except that, marvellous as such a deep and narrow hole reaching straight down two or three miles, but only a few yards across, would seem to the observer taking such soundings, it would be easy to explain, compared with the sidereal pheno- menon that Herschel had before him. We can imagine causes for a deep vertical hole in the earth's crust, but we can neither imagine any cause for a straight star-strewn projection of the galaxy in a direction exactly from the sun, nor admit the possibility that such a projection could continue if it had ever existed. That there should be several such projections would be simply impossible, even if we admitted the possibility of the existence of one. But this result, which thus conclusively proved that the principle of the new method of star-gauging was unsound, established nevertheless a most interesting fact. Since the clustering regions that yielded in part to Herschel's weakest telescopes, but not wholly even to his most powerful instruments, could not possibly be long, straight projections similarly con- stituted throughout their length, it follows that they must be clustering aggregations presenting a wide variety of stellar texture. There must be larger stars separated by wide intervals, stars not so large and separated by intervals not so wide, and stars smaller and smaller in real size and set more and more closely, till even with Herschel's most powerful telescope they could not be separately discerned. In other words, instead of penetrating more and more deeply into space, as he supposed, he was in reality scrutinizing more and more closely the stellar structure of one and the same region of space. This variety of feature within clustering regions of the Milky Way would have appeared strange to Herschel (in fact, the idea scarcely presented itself 44 OTHER SUNS THAN OURS. to him), but in our time it appears the most natural thing in the world. The analogy of the solar system, as known to us, suggests precisely such variety of structure in the greater system that Herschel was studying. Analyzed by optical powers varying in range from unaided vision to the keenest telescopic scrutiny yet available, the solar system presents a constant increase of complexity. The eyes see sun, moon, and a few planets ; the telescope reveals more planets, some really as large as Uranus and Neptune, but faint through vastness of distance ; others nearer than Saturn and Jupiter, but looking faint because small ; and yet others associated with the larger planets as dependent orbs ; more and more bodies come into view with closer and closer scrutiny of the solar domain ; yet portions still remain unresolved, such as the Zodiacal region, where astronomers more than suspect that millions of millions of nerolites and meteorites are travelling around the central orb. With this knowledge for our guidance, it seems as strange to the thoughtful student of the heavens in our time to regard the stellar system as generally uniform throughout in texture, as the diversity of texture that we recog- nize in the solar system would have appeared to Herschel. Observations of star-clouds regarded by Herschel as external galaxies, should have led him (and doubtless would in earlier years) to a similar conclusion. It is true that in many of these systems there is an apparent uniformity of stellar texture consistent with the idea that they are formed of stars of about the same size, and strewn with general uniformity through the whole region occupied by the star-cloud. Most probably, indeed, the considera- tion of these features encouraged Herschel in the belief that our own galaxy is similarly uniform in texture. Moreover, in comparing one star-cloud WILLIAM HERSCHEL'S STAR SURVEYS. 45 with another, Herschel was not necessarily led to recognize the possibility that, even as one star differs from another in glory, so the nebulae may differ much from one another in structure, regarding them for a moment as he did, that is, as external galaxies. But there was a simpler yet absolutely fatal objection, in the results that he obtained, to the theory that ran through all his work at this time, viz., that not only is the texture of our own galaxy uniform throughout the extent of the stellar system, but the same sort of star-texture exists, with con- siderable general uniformity, among all the island universes within our ken. Herbert Spencer was the first to note this objection ; but it occurred inde- pendently to me (it is, indeed, obvious) in 1867, when I had not as yet read a line of his works. That it did not occur to Herschel himself, shows . clearly how unready, in his extreme old age, he had become to analyze his results as he had in earlier years. Herschel had found parts of our galaxy un- fathomable, which showed that, in accordance with his assumptions, the outermost extensions of the galaxy are beyond the resolving power of his mightiest telescope. But the nebulae, if they are external galaxies, must lie hundreds of times farther away than the outermost parts of our own galaxy. For each one of them, from its observed size, is known to lie at a distance exceeding hundreds of times its own diameter — that is, the diameter of our galaxy, on the assumption that galaxies are all of about the same size. Thus, then, we have this absurd result, that, whereas parts of our uniformly textured galaxy, at a distance of half its diameter, are irresolvable by the most powerful of Herschel's telescopes, many similar galaxies, hundreds of times farther away — corresponding to the diminution of 46 OTHER SUNS THAN OURS. their light tens of thousands of times— are resolvable with telescopes of much smaller power ! Manifestly the principle of the second gauging method fails here again for the third time, and most hopelessly. Whether the star-clouds are external galaxies or not, the principle that Herschel had adopted for their interpretation, and in order to bring them into comparison with our own stellar system, must be given up. But we know now — I venture to speak of it as certain, though many suppose it to be but a theory of my own — that the nebulae are part and parcel of our own galaxy. Herschel's results went far to prove this, and had he but analyzed them he would have seen as much. Not only does our galaxy differ greatly in texture in its various parts, but it is as varied even in constitution as our solar system, or, rather, it is doubtless infinitely more varied in reality, though presenting to us the evidence of only about the same degree of variety. As in the solar system there are large planets and small ones, so in the stellar system there are stars of many orders of real size ; as in the former we have streams of tiny bodies, like the asteroids, so in the galaxy we find streams of small stars, as in the Milky Way ; as in the solar domain there are meteor-clouds and comets partly or wholly gaseous in structure, so in the great galaxy to which our sun belongs there are clouds of star-dust and mighty masses of nebulous matter (chiefly gaseous), like the Orion nebula. I may hereafter give a brief sketch here of the evidence respecting the architecture of the stellar heavens already obtained by astronomers. In such a sketch the work of the Herschels would hold a prominent place. I may also show the methods of survey that commend themselves for future employ- ment. My present object has been, first, to show how entirely distinct were the two methods of star- WILLIAM HERSCHEL'S STAR SURVEYS. 47 gauging that many who suppose they know some- thing of Herschel's work have hopelessly confounded together ; secondly, to point out how thoroughly the application of each disproved the assumption on which either had been based ; and lastly, to show how, nevertheless, the results obtained by each method threw useful light on the great problem that Sir William Herschel, first of all men, successfully attacked by observational niethods. CHAPTER IV. NEEDED STAR SURVEYS. In the preceding chapter on Sir W. Herschel's two methods of gauging the star-depths, I showed that, in a sense, both methods failed, one obviously to himself, the other as tested by his own method of reasoning. But let us consider what we mean when we say that either method failed, and then note what each method showed, what other methods are sug- gested by the results of applying those, and lastly, what further plans are available for the survey of the star-depths. Herschel's first method of gauging the heavens was based on the assumption that the greater the number of stars seen with a given telescope in one and the same direction, the greater the extent of the sidereal universe in that direction. It can only be said to have failed in this respect, that it showed the incorrectness of the assumption on which it was based. Herschel found that a great increase in the number of stars seen in particular directions may arise — and in many cases certainly does arise — from the clustering of great numbers of stars in their particular regions of space — a condition of things of which his preliminary assumption had taken no account. But while this involved the utter failure of the process of measurement which he had proposed to apply to the stellar universe, it by no means implied the failure of his observations to reveal any new 48 NEEDED STAR SURVEYS. 49 truth. On the contrary, the very circumstance that he had to give up his preconceived idea of stellar distribution shows that a quite unexpected discovery had rewarded his star-gauging labours. He had been able to demonstrate the clustering of stars in par- ticular regions of space, and therein lay a discovery of extreme interest. Herschel's second method of gauging the heavens was based on the assumption that the greater the telescopic power required for the resolution of the milky light of the galaxy into discrete stars, the greater the extent of the sidereal universe in the direction thus explored. This method also failed ; but it only failed in this sense, that it showed the assumption Herschel had thus made to be incorrect. In some regions of small extent he found the reso- lution of the milky light to begin with his lowest powers and continue until his highest were used, milky light even then still remaining unresolved. And although he did not himself note the point, it is manifest that this, if his original assumption had been sound, would have signified the existence of long spike-shaped projections of stars from the sidereal system, all these projections, by an incredible chance, being directed exactly from the solar system. As such a supposition cannot be accepted for an instant, it is manifest (though Herschel, then in extreme old age, did not notice this), that there must be a clustering of stars of many orders of real mag- nitude within particular regions of space — a condition of things of which Herschel's second preliminary assumption had taken no account. But here, also, while the complete failure of this second process of measurement was involved, this failure by no means implied the failure of Herschel's observations to reveal new truth. Here, as in the other case, the very circumstance that a certain idea of stellar distribution had to be given up, showed 50 OTHER SUNS THAN OURS. that a discovery of importance had rewarded Her- schel's labours. He had been able to show that the clusterings of stars already demonstrated was not a clustering of stars nearly equal in magnitude, but of stars differing enormously in real size. Some of the rounded clusters thus examined by Herschel are so limited in extent that, assigning to them a roughly rounded real form (inferable from their obviously rounded apparent form), we see that the farthest parts of these clusters are not farther away than the nearer parts in greater degree than as lOO is greater than 99. But within these narrow limits of real distance Herschel found differences of stellar size and resolvability, through some eighteen star magnitudes, which would have corresponded (had his assumption been true) to distances differing much more than a hundred differs from unity. The dis- covery that within rounded regions of the stellar universe there may exist so many orders of suns, the largest exceeding the smallest thousands of times in volume, was of extreme interest, and threw an entirely new light on the architecture of the sidereal system. In like manner it is to be noticed that Herschel's observations of star-clouds or nebulae, although by no means to be interpreted as he had at first sup- posed, are most important in their bearing on our ideas respecting the structure of the sidereal system. He regarded the nebulae as outlying universes re- sembling our own galaxy, — a grand idea justifying what is said on his tombstone, that he had broken through the bounds of our heavens — Ccelorum fierrupit claustra. It is, however, certain in reality that all these star-clouds are within the limits of our sidereal universe. Herschel's own principle of interpreting his observations, though inadequate and inexact, suffices to prove so much as this. It is certain that with his most powerful telescope he was unable to NEEDED STAR SURVEYS. 5 1 reach the limits of our galaxy ; it is manifest, there- fore, that he could not see with them the individual stars, or even the milky light, of galaxies lying far beyond those limits. Therefore all the nebulae observed by him were within these limits. There is no possibility of escaping this conclusion, unless we admit the possibility that there exist outside our galaxy others consisting of enormously larger and more brilliant stars — stars thousands of times larger than Sirius and Vega, which are themselves at least a thousand times larger than our great and glorious sun. Of course, all these results may be said to have been proved at one stroke by Sir John Herschel's observations of the Nebeculce or Magellanic Clouds, He found in those rounded regions (i.) immense numbers of stars, indicating enormous range in distance if his father's first gauging principle is accepted ; (ii.), immense varieties in the size of stars (from the seventh to stars so faint that in his powerful gauging telescope they could not be indivi- dually seen), and this, according to his father's second gauging principle, indicated also enormous range in distance ; and (iii.) great numbers of nebulae of all orders, indicating, if his father's views about nebulae .were sound, that beyond each of the Magellanic Clouds, but separated from them by enormous dis- tances if unoccupied, lie immense numbers of galaxies of suns forming two systems of such galaxies so situate, by an incredible chance, that they seem to correspond exactly in shape and position with the Nubeculae. All this is certain, if the older methods of interpretation are insisted upon. " The access to the Nubeculae on all sides," says Sir John Herschel, " is through a desert." " Intensely poor and barren regions," he says of those spaces which surround these Magellanic Clouds. In directions, then, all round the Nubecula the limits of the galaxy are very near — according to both the gauging principles of Sir W. E 2 52 OTHER SUNS THAN OURS. Herschel. Again, the Nubeculae are absolutely richer than any other parts of the heavens in nebulae, which here form apparently two clouds of star-clouds, co- inciding in extent with the Magellanic Clouds. If these nebulae are outlying stellar universes, they must lie at enormous distances beyond the Nubeculae, yet appear to be coincident in shape and position with them — a thing as incredible as that clouds in the sky should have precisely the same shapes and positions as dust marks on the pane of glass through which an observer sees those clouds. Sir John Herschel naturally rejected the idea that the features of the Magellanic Clouds are to be thus explained. Even if we could suppose, he said, that one of these clouds had the shape of a long cylinder whose axis was directed exactly towards our solar system, it would be impossible to suppose that the other is to be similarly interpreted, that a similar strange chance had set a second long cylinder of stars in space with its axis bearing exactly on the sun's family. He does not seem to notice the yet more fatal objection that each Magellanic Cloud would have to be something more than a cylinder of stars so situate ; at an immense distance beyond each cylinder, and exactly in the prolongation of its axis, there would have to be a cloud of star-clouds, to account for the multitudinous nebulae within the Nubeculae. But Sir John Herschel pointed out ob- jections enough to convince every one that the Magel- lanic Clouds have in reality the rounded form which they appear to have. Then he went on to show that, this being so, the distance of the remotest object in either Nubecula does not exceed the dis- tance of the nearest more than ten exceeds nine. Within these narrow limits of distance, he says, lie all orders of stars from the seventh down to the faintest visible in the great gauging telescope — nay, even to milky light completely irresolvable into NEEDED STAR SURVEYS. 53 stars, besides all orders of nebulae, clustering, irregu- lar, round, elliptical, resolvable and irresolvable, bright and faint All this shows that science had been quite mis- taken in supposing that the stellar universe consists merely of stars, not differing greatly in size, or much more richly strewn in some parts than in others. Just as the old idea of the solar system formed by Copernicus, as a central body circled round by six planets, has long since had to give way to the diver- sified system recognized by the astronomy of to-day, with its sun and giant planets, terrestrial planets and asteroids, large moons themselves like worlds and small moons like those of Mars, the ring system of Saturn, and finally (at present at least) the Cometic and Meteoric systems, so has the old and simple idea of the stellar galaxy had to give place to the conception of a most complex system, with giant suns, suns like our own, and minor suns, double, triple, and multiple suns, clustering aggregations, streams, branches, clouds, and complex groupings of stars of all orders, with nebulae of all kinds, stellar and gaseous, round, oval, ring-shaped, spiral, and irregular. And here I would pause for a moment to correct an idea which has been very frequently suggested in terms implying that it is the obvious explanation of what we see instead of being absolutely inadmissible. In almost all works of astronomy, when the varying degree of resolvability within cloudlike regions of star space has been mentioned, we find the minute points of light, recognized when resolution is effected, treated as if of necessity they were suns like our own, each girt round by its family of worlds. But this is altogether incorrect. It is absolutely certain that stars strewn through space like our sun and his fellow-suns (the individual stars of our constella- tions), could never appear as a milky, unresolved S4 OTHER SUNS THAN OURS. nebulosity ; for the simple reason that with in- crease of distance the individual stars, even were they as large as Sirius, would disappear long before they drew close enough together to present the appearance of irresolvable cloud. This is easily shown : — Suppose the stars visible to the naked eye to be all suns like our own, the faintest being therefore about a hundred times farther away than the brightest. Then for that spherical region of space to be removed to so great a distance that the whole set of some 6,000 stars formed a cloud as large as the moon, the centre (our sun suppose) would have to be removed to a distance exceeding more than a hundredfold the entire diameter of the sphere, and therefore exceeding more than two hundredfold the distance of the faintest visible star from our solar system. All those 6,000 stars then would lie not only enormously beyond our unaided vision, but beyond the range of telescopes of considerable light-gathering power. For, removing the faintest to a distance two hundred times greater, would cor- respond to reducing its light to one-eight millionth part of its present amount. But six thousand stars strewn over such a portion of the heavens as the moon covers would be easily separated ; the average distance between them would be twenty seconds of arc, and double stars separated by such a distance as that are considered quite " coarse." Thus, long before the individual stars were merged into each other by the effect of distance, each would be separ- ately undiscernible. Now, it should hardly be necessary to point out that to speak of stars separ- ately invisible, but lying at distances easily discer- nible (as such), forming a milky light, however faint, is utterly absurd. It is essential for the production of such milky light as we see in the galaxy, that the apparent distances of the separate stars should be NEEDED STAR SURVEYS. 55 lost through effect of distance before the stars cease to be visible. As the point considered in the last paragraph is of great importance, and very little understood (or noticed, if understood), I give the following illustra- tive tests : — Fig. s. Here we have three groups of dots all of the same size, but less closely set in 2 than in I, and still less closely set in 3. Now if the page be set up where the light falls well upon it, and then the observer retreats gradually from it, he will find that at a cer- tain distance 1 assumes the appearance of a darkish grey square, while the separate spots in 2 and 3 remain still visible. Farther away i remains as a darkish grey square, of just the same tint as before, but smaller ; 2 appears as a light grey square, and the separate spots of 3 are scarcely discernible. But passing farther away the separate spots in 3 disappear altogether from view without having coalesced, as those of i and 2 successively did, so as to form a tint. When last seen they are still at recognizable distance from each other. Moreover, the tints of i and 2 remain unchanged as you pass farther away. All that happens with these squares 5 6 OTHER SUNS THAN OURS is that they appear to become smaller with increase of distance. These illustrative tests show that the mere visi- bility of milky nebulosity in the star-depths tells something about the distribution and nature of the stars within the region observed. Stars separated by considerable distances can never appear like a diffused cloud. Stars of the same size, but some what more closely set, will appear as a very faint nebulosity if far enough away ; if still more closely set, such stars will appear as a brighter nebulosity, even when at a more moderate distance ; and a number of such stars very closely set indeed will appear as a very bright nebulosity even at a small (relative) distance. But wheresoever set beyond the distance at which nebulosity results, a star-cluster will appear neither brighter nor fainter, only larger when nearer and smaller when farther away. Thus when we see a bright milky nebulosity in a rounded region (the shape showing that we have not to do with enormously long ranges of stars in the direction of the line of sight), we know that we have before us closely-set stars, not stars strewn like those which form our constellations. We know that the stars in Cassiopeia, for instance, could never form a nebulous group such as the Pleiades appears to weak eyesight, or the beautiful cluster in the sword- hand of Perseus to the keenest vision. For we know that long before the stars in Cassiopeia had ap- proached near enough to each other — through the recession of the group — to coalesce, they would have disappeared wholly from view. Nor would any further increase of distance and the use of the telescope make any difference : the telescope would increase the brightness of the stars themselves as it increased the apparent distance between them ; and at whatever distance the stars disappeared to tele- scopic vision they would still be as far from coalescing NEEDED STAR SURVEYS. 5/ as when similarly disappearing to ordinary vision.* So that the group of stars forming the Pleiades is altogether differently arranged from the group of as many stars forming Cassiopeia ; the group forming the Beehive (Prsesepe in Cancer) is again differently arranged ; the group in Perseus differently arranged still : and in fine each star-grouping is unlike its fellows, just as the family of giant planets is unlike the family of terrestrial planets, and that family again unlike the zone of asteroids. The arrange- ments of stars are as varied as are the stars them- selves unlike in size and glory : the architecture of the stellar universe is as diverse as its materials. Suppose we could pay a visit to the midst of the Pleiades. What should we find .' According to ordinary ideas we should find simply a number of suns, each, like our own, the centre of a system of worlds. Yet it is demonstrable — and easily — that we should see around us something entirely unlike the star-strewn heavens that we now see. Probably most of the stars now visible to us would still be in sight, and scattered with much the same relation between the lustre and average apparent distance as in our present skies. But how about those stars which belong to the group we are visiting } With what lustre would the six stars shine which ordinary eyesight recognizes in the Pleiades, or the fourteen stars which some keen eyes can discern in the group ? It is certain, from the apparent size of the group, that the entire space of the Pleiades cannot be more than the fiftieth part of the distance separating the Pleiades from us. Therefore set in the middle of that * Perhaps I should rather say to ordinary vision corrected by a glass just making the stars neat and well defined ; for there is not one man in a thousand whose view of a star group is not to some degree improved by the use of an eye-glass just adapted to correct the defects of his vision — defects scarcely noticeable otherwise. 58 OTHER SUNS THAN OURS. group we should be within less than one-hundredth of our present distance from all the stars of the group. Alcyone now shines as a third-magnitude star, five others of the group as stars of the fourth magnitude. How would they appear if we dimin- ished our distance to one-hundredth part of its present amount .? Their lustre would be increased, not a hundredfold, but one hundred times a hundredfold, or ten thousand times at least. Many of them would be far more greatly increased in brightness. They would no longer be stars, but suns, just as Sirius in the great reflector of Lord Rosse, though still but a mere point in apparent size, shines like a young sun. The scene presented by the hundred stars of the Pleiades would be indescribably beautiful. In the background would lie a star scene as beautiful as the heavens we now see ; but it would be scarce notice- able amid the splendour of a hundred suns, the least outshining Sirius a hundredfold in splendour. And among these, the greater glories of the night skies within the Pleiades, there would be varieties in glory as great as among the stars of our own skies ; for the stars which seem so unequal in the Pleiades are really as unequal as they seem, since the whole group must be regarded as practically at the same distance from our earth. But this is nothing compared with what we should find if we could visit some of those glorious clusters which have been poetically described by Tennyson as " bee-like swarms of suns." The idea that the stars of those clusters are distributed like the suns of our firmament, that we have merely to count their number and say there are so many suns each girt round by its family of worlds, and each repeating not only the glory of our own sun but all the won- ders of the solar system, is demonstrably incorrect. From the observed size of those clusters we know that the entire span is less than the thousandth part NEEDED STAR SURVEYS. 59 of the distance separating us from them. Yet at the distance at which they He we can discern separately thousands of stars. Thus the average distance from star to star within some of these groups cannot be one-miUionth part of the distances separating those groups from us. The splendour of some among those stars, as seen from the midst of those groups, cannot therefore but exceed a billion* times their lustre as seen by us. Since, then, these stars can in the case of the more magnificent clusters be seen with very small telescopes, it follows that from the middle of those clusters they must shine with a glory comparable with that of our own sun, whose lustre at the distance of the nearest of our stars would not be reduced to more than one 50,000 millionth part If, then, there are worlds circling round those suns, there can then be no night for their inhabitants, but probably a constant daylight exceeding many times in glory the most resplendent of our summer days. I should be disposed for my own part to imagine rather, though I must confess I know nothing on the subject, that if there be inhabited worlds at all in connection with a glorious cluster of this sort, they must be worlds circling round the whole cluster, not round the individual stars composing it. But more probably, I should say, there are no such worlds, but those clusters will hereafter aggregate into suns which in due course will become the centres of solar systems, more or less (probably very little) like the family of planets to which our earth belongs. These considerations may serve to show what in- terest surrounds the inquiry into the architecture of the sidereal heavens. If two varieties of stellar * By a billion I piean a million millions, by a trillion a million million millions, and so forth — ^the English way of reckoning, by which a biUion, a trillion, a quadrillion, and so forth, means a million raised to the second power, the third power, the fourth power,. and so forth. The other system seems unreasonable. 6o OTHER SUNS THAN OURS. arrangement alone suggest such diversity of condi- tion, what might we not expect to follow from the consideration of all the peculiarities of stellar distri- bution which may be recognized when the heavens are carefully surveyed ? It is to such work as is thus suggested that I referred in the opening remarks of the present chapter. The failure of the two methods of gauging, devised by Sir W. Herschel, should by no means discourage astronomers from prosecuting diligent researches into the noble problems dealt with by him. Not only, as I have shown, did each failure involve an important and quite unexpected discovery, but both these failures helped to show along what lines the inquiry may best be prosecuted. I propose now briefly to consider what these lines are, touching somewhat on the work I have myself done in pursuance of this special inquiry, and deal- ing somewhat more fully with the work which has to be done (even in the earliest stages of the inquiry), wherein also I hope to bear a part. Wilhelm Struve was led by his study of the papers of Sir W. Herschel to recognize — but only indis- tinctly — the importance of combining the principles which underlie the two methods of star-gauging : — In any true survey of the star-depths it is mani- festly essential that the system of counting stars with the same telescope, and as nearly as possible under the same conditions, should be carefully ap- plied. And we shall not be led astray by this system if we do not interpret our results on an incorrect principle, as Sir William Herschel did in the begin- ning of his work. Moreover, we can apply this system in ways which at first he would have rejected as useless. He supposed that the great gauging telescope which he applied reached in all directions to the very limits of the sidereal universe, and it is clear that nothing short of such a power as he thus supposed himself to be applying could have served NEEDED STAR SURVEYS. 6 1 his purpose if our galaxy were such a system as he imagined. But that particular space-penetrating power, though it did not do what Herschel had ex- pected (because the stellar universe is not what he supposed it), and though it was unequal to the task of resolving all parts of the stellar heavens, disclosed, as we have seen, important truths. It is manifest that less telescopic power would have also given important results — seeing that the condition Herschel had supposed essential to the validity of his survey had no real existence. Nor can one see any reason to limit the diminution of telescopic power by which useful results might be obtained. Without telescopic aid at all, the distribution of stars numeri- cally might be well worth studying. Nay, it might be worth while to examine the distribution of stars visible with less than ordinary powers of vision. It was the recognition of this (possibly a half- unconscious recognition) which seems to have sug- gested Herschel's second method of star-gauging. In this he took only a very small region for survey, and examined that with constantly increasing tele- scopic power, under the idea that he was thus pene- trating more and more deeply into space. Now this kind of research, too, is manifestly essential in any true survey of the star-depths. Nor shall we be led astray by this system unless we misunderstand what we are doing. We may be penetrating more deeply into space as we increase our telescopic power, or we may simply be analyzing more and more scrutiniz- ingly a particular region of stellar space : more probably we are doing both. But if we keep our minds free from any bias one way or the other, our results will always be available for the increase of our knowledge so soon as we can co-ordinate them pro- perly together, and combine them duly with results otherwise obtained. But manifestly we must for 62 OTHER SUNS THAN OURS. this purpose extend this method of survey to larger regions than Sir William Herschel dealt with. If his principle of interpretation had been sound, his plan of applying the method would have been all that was needed. But so soon as we recognize the unsoundness of the principle, and note how that un- soundness was shown by the study of small regions of the heavens, and how important in itself was the discovery thus made, we see that results of great im- portance may be obtained frpm extending the survey by this method. Nor can we see any reason to limit the extent of the survey thus made. It may be ap- plied to the whole heavens, if only a large enough array of labourers can be persuaded to take part in the work. The study of the proper way of applying each method points, then, to one and the same result — ' viz., that the whole star sphere requires to be surveyed with every order of visual power (separately) from the unaided vision, or even from visual powers lower than ours ordinarily are, to the highest telescopic power that can be obtained. A colossal work truly : but then, fortunately, it is not necessary that the whole work should be under- taken at once. Any part of it — the survey of any portion of the heavens with such and such telescopic powers, or the survey of the whole heavens or any part of them with any definite telescopic power — means so much added to our knowledge of the archi- tecture of the complex system of stars of many orders, star-clusterings, star-clouds, and other forms of matter, which we call the galaxy. The elder Struve, recognizing the importance of combining both systems of survey, began the task by a piece of work which cannot but be regarded as very rough indeed, though it has been enthusiasti- cally admired by the late Prof. Nichol of Glasgow, and some others, who seem to have no idea of the NEEDED STAR SURVEYS. 63 real sort of work which is required in dealing with the architecture of the heavens. Struve saw that some importance attaches to the inquiry whether stars of the brighter orders — such, for instance, as can be seen with a two-inch telescope — are more richly strewn on the Milky Way than elsewhere. Clearly, according to Sir W. Herschel's earlier ideas they ought not to be. The range of distance around our sun within which such stars are included, on the assumption of generally equal dis- tribution, falls well within the breadth of Herschel's flat-disc galaxy, and (on that assumption) it is only when we pass far beyond such distances, that we come either on the vacant space bordering the flat sides of our galaxy, or on the mighty vistas of stars along its regions of greatest extension which produce the soft light of the Milky Way. Here, then, was a general test for the validity of the method which Herschel had found to fail him only in specific in- stances. The proper way of applying the test by those statistical methods which Struve loved, but which I reject as utterly inadequate for any but the simpler problems of stellar distribution, would clearly have been to have counted the stars in the Milky Way (or on a galactic zone of such and such breadth), noting the area of the region thus dealt with, then to have counted the stars in the regions of the heavenly sphere outside the Milky Way (or that galactic zone), noting the area thus dealt with, and then to have compared the wealth of stars in these two areas — ■ the galactic and the non-galactic. Supposing the charts or catalogues used for this work to have resulted from a fairly uniform survey of the heavens, or were it even only of the northern hemisphere, with a telescope of small power, the result Struve would have thus obtained would have been satisfaCf tory enough. 64 OTHER SUNS THAN OURS. This was not what he did. Probably he had not time. The process he actually applied indicated cer- tainly that he was somewhat pressed for time, since it could not have taken him much more than ten minutes. He took a certain catalogue of stars down to the eighth magnitude. In these the stars to a certain distance on either side of the equator were arranged in the order of the twenty-four hours round the celestial sphere (technically, in order of their Right Ascension) and numbered from first to last. The number of stars in the first hour was thus indicated in the catalogue as the number of the last star within that hour, the number in the second hour was obtained by subtracting the number of the last star in the first hour from the number of the last star in the second hour ; and so on for the num- bers of stars in the third, fourth, fifth hours, and so on, up to the twenty- fourth. Twenty-three subtrac- tions gave Struve all the statistics he employed. He found that the fifth, sixth, and seventh hours on one side of his zone, and the eleventh, twelfth and thir- teenth on the other, were richer in stars than the rest, in such degree as to show that the Milky Way, which crosses the equatorial zone aslant at those hours is richer than the non-galactic parts of the heavens. But it is hardly necessary to say that the real relative star-wealths of the Milky Way and of parts outside it could not be properly indicated by so rough an inquiry as this. It afforded an independent proof of the general law which Sir W. Herschel had already recognized at the beginning of the present century ; but it scarcely added more to our knowledge. My own inquiry into this particular point involved rather more labour. I proposed at first to use the catalogues and charts of Argelander, in which 324,198 stars (down to magnitude 9-10) are included, taking the numerical distribution over the Milky Way, and then over regions outside of it. But a NEEDED STAR SURVEYS. 65 few tests showed me that while this method would involve almost as much work as a process of actual charting, it would be much less satisfactory. I deter- mined, therefore, after consulting the venerable Sir John Herschel on the subject (this was but a year before his death), to chart every single star of the 324,198 in its proper place, on an equal-surface projection of the northern hemisphere, — that is, a projection in which equal surfaces on the heavens were represented by equal areas in the map. I laid down in pencil a series of radial lines a degree apart (360 in all), and ninety-two concentric circles at one-degree distances (gradually diminishing out- wards), corresponding to the particular projection which I was employing. Then in the 33,000 spaces thus formed I marked in the stars shown in the cor- responding 33,000 spaces of Argelander's forty charts. Thus I had, charted on a uniform scale, all the stars observed by Argelander and his assistants, during seven years,' in their survey of the heavens from the north pole to a distance of ninety-two degrees all round, or to two degrees south of the equator. The work occupied me in all almost exactly four hundred hours. But the result was, I think, well worth the trouble. In the first place, I note a peculiarity in the large chart of 324,198 stars, which attracts attention at once, yet is manifestly accidental, or due, rather, to the method in which the original series of forty charts, and the single chart itself, were formed. The peculiarity is a defect, though of little importance, — yet interesting as illustrating the points which have to be attended to in such work. The circular chart seems to show in. places multitudes of concentric streaks produced by the aggregation of stars along certain very narrow zones, concentric with the boun- dary of the map — that is to say, having the north pole of the heavens as their centre. As my friend 66 OTHER SUNS THAN OURS, Professor Young pointed out, there cannot conceiv- ably be any real tendency in the stars to form cir- cular zones around the pole as centre, or along declination parallels : yet such a tendency seems manifestly suggested by the appearance of the great chart when closely studied. So far as the broad results sought and obtained are concerned, this peculiarity is of no more weight than the direction of the linear streaks by which in an engraving effects of light and shade are produced. Still until or unless the peculi- arity is explained, it detracts somewhat from the con- fidence with which those broader results are accepted. Not really existing in the heavens, how does this peculiarity of star-distribution come to appear in the chart .'' The answer, though not at first view obvious, is simple enough. The wonder would be if the peculiarity had not shown itself. Argelander and his assistants, in their survey of the northern heavens, swept the skies in circles round the north pole, after the manner of survey with the equatorial telescope, which works in that sort (its main axis being directed polewards). Now herein is at once a possible cause of circular striation in the resulting charts, from the circumstance that one sweep might be made when the air was exceptionally clear, when moonlight was wanting, and other conditions for showing faint stars favourable (amongst other causes, difference of observing power among the six who took part in the work must be taken into account), while the next sweep might be made under condi- tions unfavourable for the work. This, however, is only 2u possible cause of circular striation, though in so long a series of observations it must inevitably have occurred at tinties, and so had certainly a share in producing the peculiarity in question. But there was also a sure and certain cause of striation. The field of view of a telescope is a circle, and in " sweep- ing " the centre of one field runs along a certain arq, NEEDED STAR SURVEYS. 6"} while the next field is taken a certain distance south of that arc (or north of it, according to the way the observer works). Say the field is half a degree in diameter, and the change north or south for succes- sive sweeps nearly as great of a degree, so that one field only overlaps by a little the field next north or south of it. Then it is clear that the chance of dis- cerning a faint star near the course along which the centre of the field sweeps, is much greater than the chance of discerning a star where the fields overlap ; for in one case a whole diameter of the field is available for search, in the other only a short arc. In sweeping, the star will not escape in one case if it be seen at any part of the comparatively long time during which that diameter is passing ; but in the other case, if it be not caught while the short arc is passing it will not be caught at all. Thus, it is absolutely certain that fewer stars will escape along or near the tracks of the centres of the sweeping fields than midway, or nearly midway, between the tracks of the centres. A concentric circular striation must necessarily result. To this must be added the probability that, however carefully I marked in my ninety-two circles, there may have been slight depar- tures from their true positions, whereby some of the zones were made slightly wider or slightly narrower than they should have been. This would make the striation more marked in some places, less marked in others, than it would otherwise have been, but, on the whole, would help to make it coarser, and therefore more obvious.* * It is interesting to notice how inevitably any peculiarity in the method of distribution in such cases is bound to show itself. I remember being very much struck by an example of this which arose when I was endeavouring to secure perfectly equable chance distribution for comparison with the unequal distribution of stars of various orders, which I regard as so important a feature of the stellar heavens. After trying various methods, I thought of the following : — I drew a square lo inches 68 OTHER SUNS THAN OURS. But, thus explained, the circular striation is of no more moment than the linear striation in engravings. The broad general results deducible from my equal-surface chart of 324,198 stars were very strik- ing. According to the older theory of William Herschel's we do not come near the boundaries of the sidereal universe with such a telescope as Arge- lander used. Except for some few exceptionally large suns at distances where ordinary suns would not be reached by such a telescope (only 2|- inches in diameter) there should be no greater number of stars in the Milky Way zone observed with so small a space-penetrating power, than elsewhere. Even in the side, and divided it into 10,000 squares by equidistant parallel lines in pencil, loo each way. Opening then a book of logarithms at random, I brought down a pencil point at random on the tables of figures, taking out the digit nearest the point. When I had obtained four digits in this way — as say 4725, I regarded the two first as showing the number of divisions I was to take along one side of the square, in this case 47, and the other two as showing the number of divisions I was to take along an adjacent side, in this case 25, — where these divisions crossed, that is where the 47th row from the left crossed the 25th from the right, was the square in the middle of which I was to set a point. So I went on until I had marked in many thousands, and, indeed, tens of thousands of points. For I got, others to help me in the work. Here surely was a method of pure chance, bound to result in equable distribution. But no. When the work had gone far enough, I rubbed out the pencil lines, and found a marked tendency to parallelism in certain bands where there were more dots than elsewhere, this parallelism showing itself on the vertical zones correspond- ing to the numbers 2, 3, ... 5, 6, .... 8, 9, 10, .. . 12, 13, ... 15, 16, . ... 18, 19, 20, . . . 22, 23, ... . 25, 26, .... 28, 29, 30, and so on, and on the horizontal bands corre- sponding to the same numbers. The light bands corresponded to the vertical and horizontal bands numbered I, ... 4, ... . 7, .... II, ... 14, .... 17, . . &c. The explanation was simple enough. The figures representing i, 4, 7, cover less space than the others, and by my method of taking out digits there was more chance of taking one of the fat figures, o, 2, 5, 6, 8, and 9, than one of these lean figures, i, 4, and 7. NEEDED STAR SURVEYS. 69 for such exceptionally large suns Herschel's theory made no allowance. But what is actually the case .■" These stars, which ought to be no richer on the Milky Way, are actually so much richer that, merely by their increase of wealth, they positively show the Milky Way on the chart almost as clearly as if it were mapped there ! It is manifest that the circumstances of the survey by no means favour such a result, but the reverse. Every one who has ever studied the star-depths knows that a star which is quite clearly seen when alone or with few others in the field of view, may be undiscernible when the whole field of view is crowded with stars as in the richer regions of the Milky Way. Many stars then were lost in Arge- lander's survey of the richer fields which would have been well seen had they been in poorer regions. Thus great as are the numbers of stars seen along the Milky Way regions in my chart, there would have been many more had the chance of catching faint stars here been as great as elsewhere. This I carefully tested. Reducing the power of my much larger telescope until it was about equal to that of Argelander's instrument, I examined certain rich galactic regions surveyed by him, and others far from the galaxy, satisfying myself by using various ways of reducing the field so as to exclude the blaze from many stars in the former case, that whereas Argelander and his assistants could have seen but few more stars in the non-galactic parts of the heavens, even though they had limited their obser- vations to the darkest and clearest nights, they could have seen half as many stars again in the Milky Way had they reduced their field of view in such sort as to avoid the blaze of multitudinous stars. Thus, marked though the increase of star wealth is in the Milky Way regions as observed by Arge- lander, the increase could have been very much 70 OTHER SUNS THAN OURS. greater if all the stars within the range of his telescope had been recorded. The inference from this increase of wealth on the Milky Way is obvious and important. We learn that just as the rounded rich regions along the Milky Way are leally round ij.e., roughly globular) regions of space, in which multitudes of stars of many orders of real size are strewn, so the streams of the Milky Way are real stream-shaped or branch- shaped regions in space in which not only the very small stars, as observed by the Herschels with their great gauging telescopes, are much more richly strewn than elsewhere, but also stars very much larger, and well within the range of very small telescopes indeed. Extending the principle yet further, I have made (and published) investigations into the distribution of stars visible to the naked eye, finding them much more numerous on the Milky Way than elsewhere, and otherwise less uniformly arranged than had been supposed. (In this work I used scissors and a delicate alance to give the areas of irregular regions of the heavens.) I even examined the distribution of stars down to the fourth magnitude only, finding it by no means uniform, and well worthy of attention. I have further applied the same method of chart- ing to the star-clouds (using Professor Cleveland Abbe's excellent statistical results* as the basis of my work), finding the clearest and most convincing evidence that the nebula; form part of our sidereal system. I have further charted the stellar proper motions, — the only possible way of recognizing any law in these movements. Such charts show that many large groups of stars have a common drift, so as manifestly to form separate systems. In the only * I had already published charts of the nebulae on a less complete plan before Prof. Abbe had obtained those results. NEEDED STAR SURVEYS, 7 1 case where one of these sets of stars has been dealt with by the spectroscopic method for determining motions of recession and approach, it has been found that (as I specially predicted in that case*) the stars all had a common motion in the direction of sight, as well as athwart that direction. This showed that the evidence given by charts of proper motions is trustworthy. But this is the merest beginning. We want surveys of the heavens made with many other powers, — as with a i J-inch telescope ; a 4-inch telescope (this I have partly tried, and I know the results of a full survey would be most valuable) ; with a 6-inch telescope ; and with a 12-inch telescope : the more interesting regions disclosed by such surveys as these being then examined with the highest telescopic powers that can be brought to bear upon them. It is essential that the southern hemisphere should be at least as carefully surveyed as the northern ; for no one who has ever looked at the southern skies can fail to recognize that they are much more variegated, and therefore much more likely to be instructive in regard to celestial architecture, than the skies north of the equator. Lastly, these surveys should be accompanied by widely extended study of proper motions ; by the application of the spectroscope to determine the constitution of stars in different parts of the heavens, and their movements of recession and of approach. I venture to predict that as this work proceeds (for I am sure it will in due time be undertaken), science will be compelled to give up more and more the idea of uniformity of structure within the stellar * The group of stars ^, y, S, e, and f of the Great Bear, with f ' a small companion, Alcor, called in country parts of England '■ Jack by the Middle Horse." 72 OTHER SUNS THAN OURS. universe, recognizing a grandeur and complexity in its architecture, a variety yet harmony in its move- ments, and a significance in its amazing vitality, akin to but of a far higher order than the corre- sponding qualities within the planetary system. CHAPTER V. PHOTOGRAPHING FIFTEEN MILLION STARS. A MAGNIFICENT suggestion has been made by French astronomers. I have already dealt some- what fully, elsewhere, with the work done by the photo- graphic eyes of science, directed towards the heavenly bodies. By the power of instantaneous vision which the photographic eye, unlike the human eye, possesses, the sun's cloud-laden surface has been delineated, despite the constant fluctuations of the air through which the sun has to be viewed. By the power of selecting special colours wherewith to work, the photographic eye has drawn the corona when no trace of that solar appendage has been visible to ordinary eyesight. The delicate features of the star-clouds have been depicted, through the power which the photographic eye possesses, of seeing more and more by long-continued gazing upon faintly luminous objects. And now it is proposed to do what assuredly no astronomer, nor any band of astronomers, could hope to effect, even if working for the whole duration of the longest life. It is pro- posed to chart in their true positions all the twenty millions or so of stars which are included in the first fifteen magnitudes, so that the astronomers of future generations may know for certain the aspect of the stellar heavens — to that vast depth, at least — towards the close of the nineteenth century. Let us see what are the conditions of the task. 74 OTHER SUNS THAN OURS. Using a telescope provided with a specially pre- pared object-glass of about 13 inches in diameter, MM. Paul and Prosper Henry have been able to take in a single hour photographic charts of spaces in the heavens extending 3 degrees in length and 2\ de- grees in breadth — say six moon-breadths by four and a half. Their actual plan has been to give in each case three exposures, with such slight displacements that each star is tripled, and so there can be no possibility of mistaking accidental dots on the plate for stars in the heaveas. (It might be well, however, if in the photographs finally prepared only single images of each star were given,, a preparatory plate with triple images serving for the correction of the one iinally prepared, which might have three hours of exposure without displacement.) Now, a space of 3 degrees by 24 degrees on the heavens, or 6| square degrees, is about 1-6, 11 2th part of the whole star-sphere. So that if twelve observatories, in different parts of the northern and southern hemi- sphere, were employed to photograph the star-sphere on one and the same plan, then at each observatory only about 510 plates would have to be made. Counting about fifty-one moonless nights of clear sky and still air, one night being given to each plate, the whole work would be completed in ten years. If the charts thus obtained could be combined in sets of four, in the manner already employed by MM. Henry, there would be 1,528 sheets, each repre- senting a portion of the heavens, extending 6 de- grees by 4^ ; but although Admiral Mouchez suggests this plan as desirable, it appears open to exception on account of change of scale near the edges of the plates. The number of stars which would probably be shown in this splendid contribution to the astronomy of the future would be about twenty millions. In a single plate {see frontispiece), obtained by MM. 3GKAPH OF THE PLEIADES, BY THE BROTHERS HeNRY. (Three Exposures each of One Hour. ) 7 face page ys- PHOTOGRAPHING FIFTEEN MILLION STARS. 75 Henry recently, nearly 5,000 stars can be counted ; and if 6,112 gave each such a number — say 6,000 times 5,000 — that would be thirty millions of stars. But the region shown in this particular plate belongs to a rich part of the Milky Way, and it has been shown by my chart of 324,000 northern stars down to the tenth magnitude, that there is a much greater density of stellar aggregation on the Milky Way long before the space-penetrating powers have been employed which the Herschels thought probably necessary to reach the regions whence the nebulous light of the Milky Way was supposed to proceed. If the photographic method were applied uniformly over the whole heavens, with a space-penetrating power reaching stars of the fifteenth magnitude in all directions, it is probable that about 20,000,000 stars would be shown. The great gauging telescopes used by the Herschels would show at the very least 100,000,000 — or rather would have shown that num- ber if it had been possible to bring every portion of the star-sphere under their survey. Fig. 6 is from the photograph of the Pleiades. A new era of stellar astronomy will open with his photographic work. The problems connected with the architecture of the heavens, hitherto dealt with by very imperfect methods, will now be discussed with all the advantage of at least a perfect system of survey. It is impossible, indeed, to overestimate the advantage of a system of charting over all the methods of statistical research which astronomers formerly employed. William Herschel in his first' method counted all the stars which one and the same telescope — a very powerful one, 1 8 in. in diameter — would show in different directions. He could only take a field of view here and a field of view there, not many hundreds in all, his son and worthy succes- sor in the work making similar observations in the southern hemisphere. No peculiarities of arrange- 76 OTHER SUNS THAN OURS. ment, nothing, in fact, but the roughest features of stellar distribution, could be recognized by such a method as this. It showed, however, how vast the number of stars forming our galaxy is, and it satis- fied Sir W. Herschel that the assumption by which he had proposed to interpret his numerical gauges was inadmissible, the stars not being strewn through- out our galaxy with any approach to uniformity. Herschel's second method, commonly confounded with his first (insomuch that one may often find even men like Arago and Humboldt mixing up in the same paragraphs the results of both methods of observation) was entirely different. He now no longer trusted to the use of the same telescope, turned in different directions, to tell him (after mere counting) the depth of our galaxy of stars in those directions ; he turned different telescopes, gradually increasing the space-penetrating power, in the same direction, to tell him, by the power required to re- solve the whole field of view into stars, the probable extension of the system in that direction. Herschel made many observations by this method, but in his advanced old age, when these observations had been gathered together, he did not recognize the absurdity of the result to which they tended, on the assump- tion he had employed. He found regions of the star-sphere, for instance, wherein stars of all orders were richly strewn, from those visible to the naked eye down to the faintest which his most powerful telescopes could show, and fainter orbs yet, whose ■ lustre could only be recognized as milky nebulosity (resulting from the combination of the light of many stars separately undiscernible). Around these rich regions were regions comparatively poverty-stricken, regions deserving the description applied by the younger Herschel to the spaces around the Magel- lanic Clouds, of which he wrote that " the access to the Nubeculse is on all sides through a desert." If PHOTOGRAPHING FIFTEEN MILLION STARS. JJ his assumption had been correct, these seeming clouds of many varied orders of stais, brought into view successively with increase of telescopic power, would be long cylindrical star-clusters, or rather spike-shaped projections of star-strewn space, hun- dreds of times longer than their thwart breadth, and chancing by some strange accident to have their axes directed exactly towards our place in the star sys- tem. Unlikely, one may almost say incredible, in a single case, this peculiarity would be utterly impossi- ble in several ; and the clouds so to be interpreted (if Herschel's assumption were retained) are many. Obviously we must reject this porcupine theory of the stellar system, with the solar system for the " pole" of all the stellar spines. We see that the rich cloud-like regions are real clouds of stars of many varied orders, and that in each case where Herschel had assumed (though only tentatively) that he was penetrating farther and farther into space, he was in reality only analyzing more and more scrutinizingly a complex cloud of stars. His position might be compared to that of an observer trying to gauge our solar system from a distance, who might naturally assume at first that the giant planets were much farther away than the sun, the terrestrial planets much farther away than the giant planets, the asteroids than the terres- trial planets, the meteorites than the asteroids, the small meteors than the meteorites, and the still smaller particles in comets' tails than meteors : such an observer, as soon as he recognized the association of all these objects into a system, would see that, in- stead of attributing the variety of aspect within the system to the variety of distance, he must regard it as due to real variety of size. The meteors which he had interpreted as millions of times more remote than the giant planets, he would now find to be in many cases close alongside of those large bodies, and, on the average, no farther away than the chief orb in 78 OTHER SUNS THAN OURS. the system, the great controlling sun. In like man- ner the faintest stars in the great clustering regions of the Milky Way are, on the average, no farther away than the leading orbs in the same star-clouds (which, be it noticed in passing, is by no means the same as saying that the fainter stars of the stellar-depths are, on the average, no farther away than the more con- spicuous). The assumptions made by the elder Herschel, though shown as his work proceeded to be mistaken, did not prevent his accumulated results from being most valuable. But the validity of statistical methods was shown to be doubtful. The assumptions of Wilhelm Struve were still more im- probable antecedently, and still more thoroughly discredited by the evidence. He took a zone of the heavens thirty degrees wide, assumed that the stars (down to the eighth magnitude) might be supposed to be first compressed along the mid-line of that zone, and then strewn out uniformly in twenty-four sectors, into which he divided the circular area enclosed by that mid-line. This naturally led toa result having no validity whatsoever. The fault of all such statistical methods is that in effect they depend on a process of averaging by which, even if the initial assumptions were trust- worthy, the significance of all the actual peculiarities of stellar architecture would be concealed. We want to have these peculiarities emphasized rather than hidden. Charting alone can do this effectually. But who can pretend to chart the whole heavens to any great depth around our solar system ? Struve used in his statistical inquiries about 70,000 stars, and I have shown in a single equal-surface chart 324,000 ; but what are they among tens of millions of stars within the range of Herschel's gauging telescopes ? That single chart required first seven years of observ- atory labour by Argelander and his assistants, then 4.00 hours of charting by myself ; yet it shows only PHOTOGRAPHING FIFTEEN MILLION STARS. 79 stars down to between the ninth and tenth magni- tudes, and even in regard to these is affected by all the variations arising from the " personality " of the different observers. The proposed photographic sur- vey would extend very much farther into surrounding space, would be far more trustworthy, and would be entirely independent of " personal equation." The idea is a magnificent one, and it may be hoped that the astronomers of all nations will help in carrying it out. CHAPTER VI. FIGURE OF THE MILKY WAY IN SPACE. Nineteen years ago* I wrote a paper called " Notes on Star-Streams," in which I discussed the relations presented by the Milky Way, looked upon as in reality a star-stream, and not the mere projec- tion on the celestial sphere of a widely extended disc of stars. I endeavoured to show that although Sir W. Herschel's view respecting our galaxy was perhaps the only one which he was justified in form- ing when prosecuting his celebrated star-gaugings, it is yet one which is far from being in accordance with the information which he himself gathered for us, and is still further opposed by the facts which Sir John Herschel observed during his survey of the southern heavens. And I dwelt in particular on the evidence which the strange convolutions of the Milky Way, its narrow necks or isthmuses, the knots or clustering aggregations upon it, and still more the circular gaps which pierce it, afford respecting its structure. These point to the conclusion that whatever the Milky Way may be, it is certainly not what Sir W. Herschel had supposed. But I was forced at that time to admit that the problem of suggesting the real configuration * "Intellectual Observer" for August, 1867. This paper was written, however, seventeen years ago. I found it along with the letters of Sir John Herschel I have elsewhere quoted. The paper and the illustrations appeared to me worth preserv- ing in connection with Sir J. Herschel's letters. 80 FIGURE OF THE MILKY WAY IN SPACE. 8 1 of our galaxy was more than I could manage. Its complexities seemed unintelligible ; though I did not wholly dismiss the hope of discovering a tolerably simple solution of the difficulties which presented themselves. " I may, perhaps," I remarked, " return on some future occasion to the consideration of the subject." When I thus wrote I was in hopes that the apparently intractable windings of the galaxy, as exhibited to us in the drawings of Sir John Herschel. would have been long ere this reduced into something like order. For I must admit that it seemed to me as though our astronomers had been wilfully increasing the difficulty of the problem by the perverse way in which they had chosen to regard it. It was well fitted to the noble genius of Sir William Herschel to take a wide view of the sidereal scheme. Indeed, standing where he did when he first attacked the problem, he had no choice but to select the more obvious and general features of the stellar scheme for his consideration. But as he progressed with his work he gradually began to modify many of the views which he had formed when the work was com- mencing. Or rather, I should perhaps say that he began to test the general principles on which he had been compelled to base his inquiries. And there are few more interesting subjects of study than the gradual progress by which our great astronomer made his way from one point to another, until towards the end of his life he seemed preparing to lay before the world views which, while the direct fruits of his earlier hypotheses, were yet altogether opposed to them. The work which the elder Herschel has thus car- ried so nearly to its completion fell into no unworthy hands. Sir John Herschel, inheriting his father's grand powers of generalization almost undiminished, possessing also a capacity for laborious and far- G 82 OTHER SUNS THAN OURS. sighted observation altogether equal to his father's, and a more thorough acquaintance with mathematical modes of reasoning, seemed capable of pushing the theories of the universe to that point which I believe they would most certainly have attained had Sir William Herschel lived a few years longer. But there was, I think, a difficulty in the way. The feeling I have when I rise from the perusal of any of those noble passages in which the younger Herschel presents or discusses the views formed by his father is, that he has been at times prevented from prose- cuting inquiries which seem opposed to the general direction of his father's researches, by a feeling — very natural and amiable — of respect for his father's work and fame. I could point to many passages which seem to me to force this view upon us, but I will content myself with noticing two singular illus- trations. Sir John Herschel is describing the configuration of the Milky Way in the southern heavens. He has occasion to speak of the striking brightness of the galaxy in the southern skies. Now it need hardly be remarked that on Sir W. Herschel's theory of the galaxy this great brightness is very difficult of ex- planation. That this is so, in fact, is proved by this, that, whereas Sir John Herschel felt that we could only explain the phenomenon naturally by supposing our sun to be nearer this part of the Milky Way, Professor Grant points out (very justly) that on Sir William Herschel's theory the phenomenon requires that the sun should be nearer to the opposite part of the Milky Way, for on this supposition alone would the number of stars towards the south be greatest. Sir John Herschel gives the obvious explanation, however ; and he seems to feel how strongly it is opposed to his father's theory, for he adds that the galaxy " on this view of the subject would come to be considered as a flat ring of immense and irregular FIGURE OF THE MILKY WAY IN SPACE. 83 breadth and thickness, within which we are eccen- trically situated nearer to the southern than to the northern part of its circuit." Yet he nowhere adopts this view. I feel certain that had the disc theory of the galaxy been due to any but Sir W. Herschel, the observation would have led the younger Herschel to adopt at once and finally the ring theory, though I believe he would soon have seen reason to modify his opinion of the ring's shape and figure. Again, Sir John Herschel is discussing the Magel- lanic clouds. He is impressed with the evidence they seem to afford of the fact that, within very moderate limits of distance, the faintest telescopic stars and nebulae of all degrees of irresolvability may be mixed up with stars of the eighth and ninth magnitude. Nay, he points out in his own lucid manner that, according to all the laws of probability, we must look on this fact as established beyond dispute. He sees also, as in the preceding instance, that this view is altogether opposed to his father's views respecting the universe. Yet he closes the discussion of the overwhelming evidence thus afforded against one of the most striking of his father's views with the simple remark that, " It might lead us to look with some doubt on conclusions which in former pages of this work have been some- what positively insisted upon." A certain fact is proved beyond all question, yet in the remaining pages of the " Outlines of Astronomy " that fact is completely ignored. Even as it is. Sir John Herschel's views respecting the galaxy are marked by a certain advance upon his father's. Although not definitely adopted, we must look on the ring theory of the Milky Way as that which the younger Herschel held in preference to the disc theory. Now, it will be noticed that wherever Sir John Herschel has occasion to refer either to the narrower G 2 84 OTHER SUNS THAN OURS. portions of the galaxy or to the branches which appear to extend from it, he always exhibits a preference for the view that these narrow star-beds are in reality the side views of widely extended star- strata. He says, indeed, in one place, speaking of a region where several branches of this sort are visible, " it is obviously more reasonable to suppose that these are sheets of stars viewed edgewise, than to imagine they are real columnar excrescences, bristling up from the general level." I think we must recognize in this peculiarity the influence of the preconceived opinion that not merely our sidereal system but all the parts of it exhibit a certain tendency to lateral extension, so that the existence of a columnar star-group, or of what I should prefer to call a star-stream, is improbable d priori. Otherwise, I confess I am unable to con- ceive how his intimate acquaintance with the principles of probabilities could have failed to enforce upon Sir John Herschel the feeling that the many long and narrow streams which he saw extend- ing from various parts of the galaxy must in most instances, if not in all, be stream-shaped. Nay, even with preconceived views rendering- the estimated chance of the existence of galactic star-streams only xixTi yet the existence o"f two such streams would have balanced that d, priori improbability, since we can hardly estimate at more than -^^ the chance of a sheet of stars being seen edgewise, and therefore the chance of two being so seen would be only -^. Now, Sir John Herschel saw many such excrescent streams. It will be seen at once that the existence of small streams extending from the galaxy goes far to prove the stream-formation of the galaxy itself. When this evidence is added to that which I adduced in my former paper, the conclusion seems to me to be altogether obvious that the apparent stream of milky Figs. 7 and 8. [ To face page Sj. FIGURE OF THE MILKY WAY IN SPACE. 8? light which we term the galaxy is in reality a stream of small stars, surrounding us on all sides. But I would go farther, and assert that the naked- eye appearance of the Milk)?- Way is sufficient evidence on which to ground the belief that there is a distinct ring of matter out yonder in space, and that this ring is not flattened, as Sir John Herschel thought, but is (roughly speaking) of nearly circular section throughout its length. I conceive that nothing save the perverse way in which astronomers have chosen to deal with the phenomenon would ever have led them to forget the evidence of their senses in this matter. Of course, if we insist on taking the average number of stars visible on a certain space of the heavens as indicating the density of the stars over that space, although it is perfectly obvious to the eye that there is a distinct and systematic arrangement of the stars there, wholly negativing our initial supposition, we must expect to be misled. With all respect for the elder Struve's labours, I must admit that this seems to me to be what he has done in his famous distribution of the stars accord- ing to zones of given galactic polar limits. Consider, however, fig. 7, which represents the galaxj' as actually seen in the heavens,* and it be- comes wholly impossible to believe that we have to deal with the projection upon the celestial sphere of a widely-extended cloven disc of stars. The view does not account for one of the peculiarities of the galaxy proper, however justly it may seem applicable to the sidereal system. The great break in Argo (opposite line i in our figure) is of itself sufficient * The mode of projection must be conceived^to be as fol- lows : — -Suppose that on a celestial globe a band is taken, including the whole of the Milky Way, and that this band is spread as a long, straight slip on a plane surface. If, then, we conceive the band turned into a circular strip, from the uniform contraction of one edge, we shall have such a map as fig. 7. 86 OTHER SUNS THAN OURS. to negative the disc theory ; so is the coal-sack near Crux, and so are the somewhat similar vacancies in Cygnus and Argo. The fact, too, that the second stream (which has led to the assumption that the sidereal disc is cloven) is not continuous, is one which cannot possibly be explained on the disc theory. But, although one may feel convinced that the galaxy is really a stream of relatively small stars surrounding our heavens, it has always seemed ta me a very difficult matter to account for the various phenomena presented by the Milky Way on any reasonable hypothesis. It was easy to see that, what- ever hypothesis we adopt, we must be prepared to admit of the existence of great irregularities. In fact, as such a stream as I conceive the Milky Way to be would be subject to a number of attrac- tions, swaying its length here in one direction, there in another, those irregularities were to be looked for independently of any considerations founded upon the observed appearance of the Milky Way. But there were certain features which I felt that any hypothesis for which support could reasonably be claimed ought to explain. The difficulty I found was in conceiving how, first, the interruptions, secondly, the variations oj brilliancy, and thirdly, the lacuna in the Milky Way, could be accounted for by any single stream how- ever shaped. An explanation, which accounted for the interruption opposite line i, left the interruption opposite line 2 unaccounted for. Again, I did not find it easy to account for the sudden access of brilliancy at 8, the extreme faintness at 7, or the fact that of the^two branches starting from 5 the fainter becomes presently the brighter, and vice versd. The three coal-sacks also were a great mystery to me. I have again and again attacked the problem (which now seems perfectly simple and easy) with- FIGURE OF THE MILKY WAY IN SPACE. 87 out being able to imagine a stream of reasonable figure which would account for these peculiari- ties. At length (more than two years after I had come to the conclusion that the galaxy is really a spirally formed ring of generally circular section), by looking on the break opposite as due to increase of distance, not to a real interruption of continuity, I was able to construct a single spiral curve which seems com- pletely to meet all the requirements of the problem. This curve is exhibited in fig. 8, which is supposed to exhibit the actual figure of the galactic spiral in space. It is so situated that the various lines drawn from our sun, supposed to be at S, intersect the various portions of the figure representing the real galactic stream, opposite the regions in which these lines meet the figure of the galaxy on our heavens. We see that line I passes through a gap between the two loops of the galactic spiral. This seems (to begin with) a simple explanation of what has hitherto been admitted to be one of the most per- plexing features of the Milky Way. Passing to position 2 the line crosses two branches of the curve, and the coal-sack is accounted for by the deviation of one branch (or both branches) slightly from the mean galactic plane. From position 3 the line crosses one branch at a very small distance, the other being much farther off. This corresponds closely with the appearance of the two branches, the continuous one being very much the brighter, and some portions along this part of its length being described by Sir John Herschel as singularly bright. It is also well worthy of notice that the two stars which are nearest to the sun (so far at least as obser- vation has yet shown) lie along this branch of the galaxy — a Centauri very nearly where the branch approaches closest to the sun, 61 Cygni in direction 88 OTHER SUNS THAN OURS. S5, where the branch is some three times farther off.* The farther branch attains, along S4, so great a distance from the sun as to become invisible. This corresponds with the mode of the discontinuity of this part of the Milky Way, for each end of the broken division loses itself, not terminating abruptly like the two fan-shaped terminals opposite the line Si. Near this portion of the circuit we are provided with an explanation of what had always been looked upon as a great difficulty. Where the two branches start from the coal-sack in Cygnus (on Ss), the northern branch is much the brighter, but presently the northern branch grows fainter and ultimately vanishes, while the southern grows brighter and brighter. This is fairly accounted for by the figure I have assigned to the .spiral. The projection at 6 may be accounted for by assuming the end of the spiral to be curved back- wards as I have shown it. Lastly, the faintness at 7, the projection at 8, and the vacuity at 9 are obviously accordant with the figure given to the end of the spiral which falls opposite the lines to these parts. Without asserting that the actual figure of the galaxy in space is that shown in fig. 8, I yet think it probable that the order of its windings resembles that shown in the figure, I believe, however, that there are many irregularities not merely in the direction in which the spiral extends through space along its general plane, but in directions inclined to that plane. The appearance presented by the Milky Way in Aquila and Scorpio is strongly suggestive of such peculiarities in the real figure of the spiral. * According to the annual parallaxes assigned to these stars, 61 Cygni is between two and three times as far from us as a. Centauri. FIGURE OF THE MILKY WAY IN SPACE. 89 I feel convinced, further, that the study of the Milky Way as presented in fig. 7 will at once dis- pose of the notion that the galaxy can be either a cloven disc or a flat ring, or that the section athwart any branch of it can be otherwise in general than roughly circular. CHAPTER VII. THE SIDEREAL SYSTEM FATHOMLESS* It is commonly supposed that Sir W. Herschel's plan of star-gauging demonstrated that the sidereal system has limits to which his gauging telescopes penetrated (save in a few directions), and that even where the system has its widest extent its limits are certainly attainable by such telescopes as men may well hope to construct. I would invite attention to certain evidence pointing to a very different con- clusion. It is perfectly clear that if the sidereal system have the figure hypothetically assigned to it by Sir W. Herschel, that of a lens-shaped stratum throughout which stars are distributed with tolerable uniformity, then we must accept the evidence adduced by him as sufficient to prove that we can attain to the limits of this stratum. To use the words of Professor Nichol: "When an eye is directed towards a pro- longed bed of stars, there is no reason to fancy that it has reached the termination of that stratum, so long as there appears behind the luminaries, which are individually seen, any milky or nebulous light, such light probably arising always from the blended rays of remoter masses. But if, after struggling * This paper, like that on the Figure of the Milky Way, was •written when I was in correspondence with Sir John Herschel about the architecture of the heavens. 90 THE SIDEREAL SYSTEM FATHOMLESS. 9I long with a nebulous ground, we obtain a telescope that gives us additional light with a perfectly black sky, we then have every reason the circumstances can furnish on behalf of the supposition that at length we have pierced through the Stratum — a probability, indeed, which can be converted into certainty in only one way, viz., when no increase of orbs follows the application of a still larger instrument." Sir John Herschel also says that in those regions where the zone is clearly resolved into stars well separated and seen projected on a black ground, it is certain if the ordinarily accepted theory be correct, that we look out beyond into space. But this conclusion would no longer follow as a necessary consequence of such observations if, instead of regarding the sidereal system as of the figure and structure suggested by Sir W. Herschel, we supposed it to consist of clustering aggregations (including streams under that expression) of stars of every variety of magnitude. Then, in struggling with a nebulous ground, we should not be penetrating farther and farther into the celestial depths, but should be simply analyzing more and more searchingly a definite aggregation of stars. Let us consider a noteworthy instance, interesting not only because it illustrates the mistakes which might arise from falsely assuming a certain uniformity in stellar aggregation, but because it shows how a thoughtful astronomer like Sir W. Herschel would instinctively recognize^ under such circumstances, the fact that he was going astray, and would be capable of quietly relinquishing views on which he had before laid considerable stress. In the constellation Perseus there is a magnificent double cluster, visible to the naked eye on tolerably clear nights, and presenting, even in small telescopes, a scene which forces sensations of awe and reverence upon the least thoughtful mind. With a large tele- 92 OTHER SUNS THAN OURS. scope the spot "appears lighted up," says Nichol, " with unnumbered orbs, and these pass on and on, through the depths of the infinite, until even to that penetrating glance they escape all scrutiny, with- drawing into regions unvisited by its power. But shall we adventure into these deeper retirements .■" Then assume an instrument of higher efficacy, and lo ! the change is only repeated ; the nearer stars now shine more brilliantly ; those scarce observed before appear as large orbs ; and behind a new series begins, again shading gradually away, leading to- wards further mysteries ! The illustrious Herschel penetrated, on one occasion, into this spot, until he found himself among depths whose light could not have reached him in much less than four thousand years. No marvel that he withdrew from the pursuit, conceiving that such abysses must be endless ! " But this conclusion, that the light from the farther- most parts of the cluster occupy some forty centuries in reaching us, while the light from the larger stars in the cluster, according to the usual estimate of star magnitudes, would occupy but one or two centuries, brought with it perplexities which Sir W. Herschel was too clear-sighted not to recognize. It required that the real shape of the cluster should be some- what as is shown in fig. 9, in which S is the sun, and S B is some twenty times as great as S A. On no other supposition could the peculiarities of the cluster be explained, so long as it was understood that a general uniformity of magnitude and distribu- tion prevails among the component stars. Sir W. Herschel was thus led to recognize the cluster as including within its bounds stars varying greatly in real magnitude. Nay, he pronounced the opinion that we have in this cluster a sort of nodule of the Milky Way — a distinct clustering aggregation of stars v/ithin the limits of the galaxy. Now let us consider what such a change of view ^ <: o o THE SIDEREAL SYSTEM FATHOMLESS. 93 meant. Instead of supposing that each increase of telescopic power had enabled the observer to pierce farther and farther into the sidereal depths in this direction, Sir W. Herschel now saw that all the successive investigations had dealt with the same region of space. The difference was as great as though an astronomer, discovering new asteroids and conceiving their minuteness to be due to distance, so that they all lay hundreds of times farther off than Jupiter, suddenly learned their true nature, and that all his researches had dealt with a zone far within the orbit of the giant planet. Are we quite sure that a similar error does not affect the whole system of star-gauging, or rather the fundamental principle on which that system is estab- lished } What real evidence have we that, when we are poring more and more searchingly into the ecesses of the great star-girdle, we are passing ever to distances farther and farther from the sphere of the lucid stars .' It seems to me that, before accepting the results which have been supposed to flow from the star- gaugings, we are bound to inquire somewhat more closely than has yet been done into the question whether the probabilities are in favour of that general uniformity of distribution and magnitude on which the plan was based. And here an important point presents itself to our consideration. Admitting this fundamental hypothesis, it is very obvious that we must pay no attention to the signs of special laws of association among the lucid stars. These stars are altogether dissociated from each other in reality, however they may seem associated — if only that hypothesis is correct. But when we are inquiring whether that hypothesis is correct, these signs of association are all-important for our guidance. We are bound to inquire whether they can be accidental. And so we 94 OTHER SUNS THAN OURS. are no longer free to smooth the star-groupings away by taking averages. This bears in a very important manner on the problem presented by the Milky Way. Sir John Herschel, following very accurately the law of star- gauging, compares the total number of lucid stars on the galactic zone with the total number on the rest of the sky, and iinds no trace of any aggregation in the former region. Hence he concludes (very justly, when once the fundamental law is accepted) that there is no real association between the lucid and the telescopic stars on the galactic zone.* But suppose that instead of considering the galactic zone, instead of spreading the galaxy over a belt which it does not really cover, we look at the galaxy itself And suppose, further, that as a first process of examination we compare the number of lucid stars falling on the galaxy with the number falling on the dark rifts and coal-sacks in the Milky Way, and on the space which separates the two branches where the galaxy is double. Doing this, we find at once the most striking evidence that the lucid stars are closely associated with the telescopic galactic stars ; for we find a marked disproportion between the number of stars on the dark regions and the area covered by these regions. In many places, especially in the southern heavens, we find the very shape of the Milky Way indicated by the stars which lie round the border of the dark regions, but * In his " Outlines of Astronomy " he uses expressions which would seem to indicate that he had forgotten the facts very clearly established and described on pages 381 and 383 of his observations made at the Cape of Good Hope. Nor can one wonder at this when one considers the wonderful range and extent of the observations recorded in that most valuable treatise, second only (if second) in value to the series of papers by his father, in the Philosophical Transactions of the Royal Society. THE SIDEREAL SYSTEM FATHOMLESS. 95 withdraw themselves, so to speak, from those vast openings into space. Take as an illustration the coal-sack in Crux. Is it an accident that over this large dark space, cover- ing about 50 square degrees, there is not a single lucid star, while all round its borders lucid stars are strewn in plenty ? The whole surface of the heavens exceeds the coal-sack some 800 times in extent ; and as there are about 6,000 lucid stars, one might expect seven or eight such stars to be found in the coal-sack. But this is far from being all. The neighbourhood of the coal-sack is much richer in lucid stars than other regions in the heavens ; so that it is just where stars should be most richly distributed that this vast black spot makes its appearance. The question whether the absence of stars from the coal-sack and their presence in great abundance in the Milky Way around that vicinity are to be regarded as a mere coincidence can scarcely be doubtful, I think, to any one who studies thought- fully the portion of the galaxy depicted in fig. 10. Nor, perhaps, is the way in which the sharply de- fined semicircular cavity on the right is associated with a semicircular stream of stars less significant. No one who examines this region thoughtfully can doubt, I should imagine, that the lucid stars seen in it are mixed up with the telescopic stars forming the Milky Way here. But if we admit that such evidence as this (and much more of the same kind might be adduced did space permit) should lead us to regard the Milky Way as forming a stream of really small stars, swayed into its present figure by the large ones in its neighbourhood, it might seem that, so far from show- ing that our sidereal system has no limits, we should have gone far to prove that its dimensions are much smaller than had been imagined. It is true that, according to these views, the small 96 OTHER SUNS THAN OURS. : stars in the. Milky Way. would be far nearer to us than has been commonly supposed. But, on the other hand, it would follow with equal certainty that we cduld: no longer imagine we had even in any one direction pierced to the limits 'of the sidereal system. If in searching into the depths of any part of the Milky Way we are, in truth, merely searching more and more closely within a definite group of stars, beyond that group there may lie at enormous dis- tances other groups which no telescope we can con- struct may even render visible. It was only, indeed, whilst it was thought that the sidereal system is continuous throughout its limits that astronomers could hope to say where those limits lie. If, on the contrary, I am right in believing that the sidereal system consists of aggregations of every conceivable form, those aggregations may extend into space, millions on millions of times beyond the limits of the most powerful instruments man may ever be able to construct. CHAPTER VIII. SUNS AND METEORS. It may seem strange to associate suns and meteors, fixed stars and shooting stars. One can scarcely imagine bodies more unlike — suns, the mightiest, because the most massive, of all the subjects of astronomical research, and meteors, many of which are so small that in their brief rush through our air they are entirely dissipated, and in a sense destroyed. For millions, nay, for hundreds of millions of years, a sun endures, pouring forth moment by moment supplies of light and heat — the life of worlds circling around him — in quantities so enormous that the human mind is utterly unable to conceive them. The falling star glows but for a few seconds, and then its brief career comes to an end. Weighed in the scales of science, the suns which people space are found to outweigh, severally, such globes as our earth, hundreds of thousands of times : the falling star has also been weighed, and its average weight is found to be but a few grains ! Yet, as shooting stars have been unmistakably associated with comets, which seem so utterly unlike them, so have they now been connected, by evidence which seems too strong to be resisted, with suns. Quite recently, indeed, meteors of a certain kind have been discovered which tell us that respecting the noblest order of suns which no instruments made by man could have revealed. 97 H gS OTHER SUNS THAN OURS. Let us briefly consider the line of reasoning by which it has been shown that large numbers of the meteoric bodies which reach our earth from outer space have been ejected from the interior of suns, or of bodies in a sunlike state. We may then examine the new discovery, and consider its bearing on the theory of the origin of meteors. In former times it was the received theory respect- ing meteors that they had their origin in the upper regions of the air. But it was at length proved that, instead of that long-received theory, a theory which had been rejected as too absurd for credence must be accepted. It was found that meteors reach our earth from interplanetary space. As Humboldt well expressed it, " They bring to the earth extra- terrestrial matter ; they are the only messengers which reach us from regions outside the world on which we live." But the nature of their paths was long unknown. All that had been proved was that they travel in flights around the sun as their ruling centre. The proof was twofold. Because shooting stars are seen in showers on special days of the year — not of each year, but still so often as to show that the coincidence of date is no mere accident — it is certain that they travel on paths crossing the track of the earth at particular points. Each star-shower having a special date forms a distinct system. The second proof was equally decisive. The meteor-paths during any great display always seem to radiate from the same fixed point on the star sphere, no matter how many hours the display may last, or how much, therefore, that point may change in position with regard to the horizon. It follows that their paths are parallel before they reach the earth. The last point is to be specially noticed. It not only affords a subsidiary proof of what was already established by the agreement of dates. It tells us SUNS AND METEORS. 99 something new about the meteors and their move- ments. The observer on earth is carried round the earth's axis during the display by the earth's motion of rotation. This m.otion, though slow compared with the movement of the earth in her revolution around the sun, is nevertheless considerable in itself At the equator, a point on the earth's surface moves rather more than 1,000 miles an hour ; in latitude 45° north or south, the rate of motion is about 750 miles an hour. (London is carried round the earth's axis at the rate of more than ten miles per minute.) Now the earth travels round the sun at the rate of i8| miles in a second, and meteors would usually cross the earth's track with velocities greater than these, since a body travelling (as most meteors travel) around the sun, on an orbit extending far beyond the earth's, would have at the earth's distance from the sun a velocity of about 26 miles per second. The effects then of the earth's rotational movement, as hour by hour an observer's direction of motion (due to this cause) is altered, can but slightly modify the apparent direction of meteoric motion. Still it might be expected that in many cases these effects (which may be compared to the apparent change in the direction of rainfall, as our motion through the rain, in walking or riding, is modified) would be recognized. The circumstance that no observer of meteors has ever detected such effects shows that in all cases hitherto dealt with the velocities with which meteors encounter, or overtake, or pass athwart the earth are enormously greater than the velocities with which points on the earth's surface are carried round her axis, and greater also than the velocities which the earth can communicate to bodies approaching her from outside.* * The velocity which the earth could communicate to a body drawn to her surface from an indefinitely great distance, by her own attraction only, would be nearly seven miles per second ; H 2 lOO OTHER SUNS THAN OURS. From Olmsted's demonstrated theory of meteors (the credit of which has been very calmly bestowed of late on persons who had done no more than note a few circumstances consistent with it) has been followed, since 1866, by a series of interesting dis- coveries. It has been shown that the meteors of November 13-14 travel in a period of 33J years round the sun in a path extending beyond the orbit of the planet Uranus, and passing very close to this orbit at one point. It has been shown further that the meteors of August lO-ii and of November 13-14 travel on the tracks of known comets. It has been rendered highly probable that every meteor system tells us of the course of a comet, though not necessarily of a comet now in existence, while every comet is followed by a train of meteoric attendants. This train, by the way, must by no means be con- founded with the comet's tail — a very different forma- tion and occupying an entirely different position^ In the only case where the earth has ever been known to be approaching the track of a known comet, prediction was made (by myself and Professor Alexander Herschel) that a display of shooting stars would be seen, radiating from a particular point of the heavens, at the time when the earth was plunging through that comet's train of meteoric attendants ; and this prediction was fulfilled to the letter. We may fairly infer that what has been shown of all the comets whose paths have crossed the earth's track is true of comets generally. When so much as this was known about shooting stars it was natural that astronomers should begin but the bodies which come near the earth, or actually encounter her, are already travelling, for the most part, with much greater velocities, communicated by solar attraction ; and she has not time, during their swift rush towards or past her, to impart more than a tithe of the 'velocity which she could communicate were she alone at work upon them, and they had no sun-im- parted velocities. SUNS AND METEORS. lOI to form ideas as to the origin of these bodies. Accordingly a theory was advanced in 1866 by Signor Schiaparelli of Milan, which, because no one at the time dwelt on any of its shortcomings, or advanced any other theory, has come to be regarded by many as an accepted theory, and was so spoken of recently by Professor Young, of Princeton, N.J., in his farewell address before the American Asso- ciation for the Advancement of Science. Schiaparelli's theory was this : — He assumed that flights of meteors are travelling about through the realms of interstellar space in the form of nebulous clouds. Under the attraction of some sun towards which their course has already in some degree directed them, they travel towards the region where- in his family are travelling. If by chance a flight of meteors should come near enough to one of the members of such a family, it is deflected from the course it had been following, and may (under par- ticular conditions) be retarded. If so the future course of that flight of meteors will be a closed though eccentric orbit around the sun attended on by that disturbing planet. Such closed orbit will necessarily pass through the point where the disturb- ance v/as produced by which the meteor flight was, in a sense, captured. The theory requires further that an immense number of such captures should be made. For our earth passes through great numbers of meteor flights, and it is certain that for each meteor system, (among those captured) through which our earth passes, there must be millions to which she does not draw near. That this speculation — for it is obviously nothing more — should be described by so careful a student of astronomy as Professor Young, in terms implying that it is a theory based on the thorough investiga- tion of an adequate amount of evidence, is strange, to say the least of it. One speaks of the Coperni- 102 OTHER SUNS THAN OURS. can system as the received theory of planetary motion, but even Laplace's widely-known hypothesis of the origin of the planetary system is not called the " received theory." Newton's theory of universal gravitation is received, but Le Sage's speculation respecting the origin of the force of attraction is regarded as a speculation only. In like manner the theory that meteor systems travel around the sun, or rather that all meteoric bodies reach our earth from outside with planetary velocities, is established by evidence which cannot be shaken ; but the sugges- tion that meteors are drawn from interstellar space by our sun's attraction, and then by the casual intervention of one or other of the giant planets forced to travel on a closed path around the sun, is but a speculation, as little based on any real evi- dence as the old-fashioned idea that rain comes down upon the earth from some great reservoir of water above the crystalline. Of the general idea that meteors, and therefore comets, come to us from interstellar space, it may be said that in one sense it is manifestly probable, if not certain, in regard at least to many systems of meteors. Of many comets and meteors we have to admit that unquestionably the region whence they came on their last visit to the earth was that vast realm outside the solar domain which we call inter- stellar space. It is one thing, however, to admit this, another and a very different matter to regard interstellar space as a sort of breeding-place for meteors and comets. To explain them thus is to interpret a marvel by a miracle. It may be difficult to say whence meteors came to occupy in such incon- ceivable numbers the interstellar spaces ; but it would be hopeless to attempt to show how they might be understood to have been there from the beginning. But while there is this overwhelming negative objection to Schiaparelli's speculation, that in effect SUNS AND METEORS. I03 it explains nothing, there is a positive objection of the most decisive nature. It is one which I pointed out long since, one whose validity has been admitted, and one which has never yet been in any way answered, though Professor Young has suggested that possibly some way of answering it may yet be suggested. I will not here enter on the considerations, chiefly mathematical, on which the objection I am about to indicate is based. I will note only what is the cer- tain result of applying mathematical tests. The giant planets cannot do what Schiaparelli's theory requires that they should do. The individual mem- bers of a flight of meteors travelling from interstellar space towards the solar system may chance to pass near enough to one of the giant planets to be caused thenceforth to travel on closed paths around the sun ; nay, the flight itself might be captured (in this sense) bodily. But there is no possible way in which a flight of meteors, consisting, like the November meteors (the Leonids*) and the August meteors (the Perseids*), of many billions of billions of discrete bodies, could be so captured by a member of the sun's family, even by the giant Jupiter himself, as to travel on the paths which these systems actually pursue. As a matter of fact, if the Leonids have been captured at all, as Schiaparelli imagined, it must have been by Uranus, whose capturing power is utterly insignificant compared with that possessed by Jupiter and Saturn ; while the Perseids, if captured by any member of the solar system, must have been captured either by some planet exterior to Neptune or by the earth herself ; for the Leonids only ap- proach the orbit of Uranus and the earth in their * The reader is not to suppose that the Leonids are the only November meteors, or the Perseids the only August meteors ; I add these names to show which particular set of November meteors and which particular set of August meteors I am referring to. I04 OTHER SUNS THAN OURS. course around the sun, while the Perseids approach the orbit of no known planet except the earth. Now, taking the Leonids (for, be it observed, a single instance will suffice, and the Leonids have long been regarded as strikingly illustrating Schiaparelli's theory), we find that for a single member of this family to have had its path changed from one pass- ing out into interstellar space to one having a period of 33^ years — the actual period in which the Leonids complete their circuit — that meteoric body must have passed very close indeed to the globe of Uranus. A certain amount of the meteor's motion would have had to be withdrawn by the attractive power of Uranus, and as the velocity eventually abstracted is only the excess of the quantity abstracted during one part of the time when the body was near Uranus over the quantity added during the rest of that time, it is clear that Uranus must work very hard to pro- duce the desired effect on a body which rushes past the planet with a sun-imparted velocity of several miles per second. When details are considered, it is found that the approach of a meteor to Uranus, as the meteor came in from outer space, would have to be so very close as to preclude the possibility that a flight of many billions of billions of meteors could all pass near enough to have that path assigned to them along which all the Leonids actually travel. And so with other cases — with every other case where the actual periods, and therefore velocities, of meteors are known. Despite the opinion of Prof. Young, that in some way or other this objection may be explained away, I venture to say with the utmost confidence (and I think undue confidence about such matters is not a fault with which I can be charged) that the giant planets cannot have captured one of the flights of meteors whose true period of revolution has been determined. It may be that some among the foui SUNS AND METEORS. I05 hundred or so of meteor systems which the earth encountei's in the course of each yearly circuit around the sun have been captured in this particular way ; but, so far as known facts are concerned, and espe- cially those known facts which led Schiaparelli to formulate his so-called theory, it is certain not only that we have no evidence in its favour, but that all the real evidence is opposed to it. It was this which led me to believe that meteors have had their origin, or that at any rate multitudes among known meteor systems have had their origin, in another way. Note, first, that the researches of Stanislas Meunier and others have led many — as Tschermak and Ball, for example — to the opinion that some at any rate among the meteors annually encountered by the earth are her own children. In other words, there are reasons for thinking that during some remote past period the earth had the power, being then full of the fiery energies of planetary youth, of ejecting from her interior flights of missiles — clouds of world-dust, so to speak — with such velocity that the matter thus ejected was free thenceforth to travel around the sun, with no other subservience to its parent orb than is involved in the circumstance that, for ever thereafter, the paths of such ejected missiles would cross, or pass very near to, the track of the earth. With regard to this idea, which at first seems fanciful in the extreme, I may remark that there seems reason to believe that every orb in space passes through stages of orb-life which may be divided roughly into three : the sunlike, the earth- like, and the moonlike ;* and therefore we must re- * Another classification may be suggested — the glowing vaporous state (like the sun's), the fiery state (like that in which the giant planets seem to be), the life-bearing state (like the earth's), the state of old age (of which Mars seems to afford an example), and the death-like state (which the moon seems to have reached). I06 OTHER SUNS THAN OURS. cognize in the past history of our earth a time when her energies were far more active than those she now has. We cannot infer her power of ejecting matter from her interior, when she was in the sunHke state, from that which she possesses now when she is in the middle of the life-bearing portion of her career. When she was a sun she was a very small sun, a mere dwarf compared with the giant Jupiter when he was a sun, and a mere speck of light compared with the mighty sun which rules our system. Yet she probably possessed then eruptive powers com- pared with which those she now possesses are as nought. Yet Krakatoa taught us recently, as at other times the earth-throes of Peru and Chili, of Sicily, Naples, Spain, and Iceland have taught us, that the earth's eruptive energies are even now in no sense contemptible. The probabilities are at least highly favourable to the theory enunciated by Tscher- mak. For the immense numbers of sporadic meteors encountered by our earth almost compel the belief that her track must be regarded as in a sense in- fested by meteors — crossed, that is, by greater num- bers of these bodies than traverse similar parts of the solar system outside, or within, or above, or below (north or south*) of the earth's path. This would mean, of course, that the earth has had some- thing to do with the strewing of this track with meteors ; and as the earth most assuredly has never had the power of drawing meteors from paths on which they had entered under solar influence (as Schiaparelli imagined that the giant planets might have done) it seems to follow inevitably that the earth has given birth to this surplus stock of earth- crossing meteors. Let it next be noticed that there are certain * It is as correct to speak of north and south with reference to the plane in which the earth travels, as with reference to the plane of the earth's equator. SUNS AND METEORS. I07 families of comets which have been for many years associated with the giant planets. Many years ago, and long before I recognized the real meaning of the phenomenon, I wrote an essay, which appeared in a weekly magazine of wide circulation, in which I treated of the " comet-families of the giant planets." I gave this name to certain families of comets which, though circling around the sun as their real attract- ing centre, yet have paths approaching so near to the orbits of the giant planets that we may fairly regard these comets as in some way or other dependent on the giant planets — each on the particular giant planet with which it thus seems associated. Now, as comets are known to be followed by trains of meteoric attendants, we may say that we have here a phenomenon closely akin to, if not practically identical with, the peculiarity in relation to the earth's orbit which Tschermak and others have endea- voured to explain (and, as I think, have successfully explained) by assuming that millions of years ago the earth herself ejected those particular meteors which form as it were the extra population of the earth's orbit region. So that we seem justified in adopting here, also, a similar explanation. Of course, if Schiaparelli's theory were anything more than a speculation, and still more if it deserved to be re- garded as a received theory, we might hesitate before we rejected what would be, in fact, an explanation of the very peculiarity we are considering. But we have seen that not only has Schiaparelli's theory no claim to be regarded as a received theory, or as a theory at all in any proper sense of the word, but the objections to it are, in fact, absolutely in- surmountable. We therefore turn to the other ex- planation as one which here naturally suggests itself — we inquire, at any rate, whether the cometic and meteoric families of the giant planets may not be regarded as originally ejected, in the form of mete- Io8 OTHER SUNS THAN OURS. oric streams, from the giant planets, when these were in the sunlike state. It is manifest that we are justified in assuming that if the earth ejected meteoric bodies when she was in the sunlike state the giant planets would have done so likewise. Therefore there are d, priori rea- sons for regarding as probable the theory to which we have thus been led by ^ posteriori considerations. Moreover, as the giant planets are still in a semi- sunlike state, we see that in all probability the meteor streams expelled from these planets would retain something of their original coherence ; that is, they would appear in company with comets (each comet representing a cloud of meteors originally ex- pelled as a coherent group). Thus we could under- stand the existence of the comet-families of the giant planets, though, of course, we can also under- stand that many comets formerly belonging to these families have disappeared as comets ; indeed, we have been able to watch, in the case of Biela's comet, the process of disintegration, by which one of the members of Jupiter's comet-family has ceased to exist as a comet, and remains only as a stream of meteors. But now two problems of interest present them- selves to our consideration. In the first place, we have in the sun an example of an orb in that parti- cular stage of orb-life during which, we have been led to suppose, meteoric ejection takes place, and we are naturally led to inquire whether the sun ever ejects flights of discrete bodies from his interior ; and this inquiry will naturally be extended to his fellow-suns the stars. In the second place, we are led to ask how those comets and meteor streams are to be explained which assuredly have not been ejected from the earth or any of the planets ; and this inquiry will have to be extended to those comets and meteoric streams which not only cannot have SUNS AND METEORS. 109 come from any member of the solar system, but cannot possibly have been derived from the central ruler of that system. Now, among the remarkable discoveries made by means of the spectroscope, one of the most striking has been the recognition of tremendous solar dis- turbances of an eruptive, or rather of an explosive nature. In 1872, Prof. Young, of Princeton,* N.J., observed a solar eruption, in which what looked like filaments of glowing hydrogen (each many thousands of miles in length !) seemed to travel upwards from the sun's surface at the rate of about 145 miles per second, till they had reached a height of not less than 210,000 miles. Even then they did not cease to ascend ; but, losing their lustre, faded out of view. If shreds of hydrogen were really shot out on that occasion we should scarcely find in the event any- thing bearing on the matter before us — the possible ejection of meteoric matter. But no one who con- siders the phenomenon with attention, or studies the evidence obtained in regard to it, can for a moment imagine that what look like ejections of glowing hydrogen can be really of that nature. It is obvious alike from d. priori considerations and d. posteriori evidence that the jet-like streams of hydrogen are in reality the tracks of ejected matter, solid or liquid. For, not only is it impossible that streams of such a substance as hydrogen should be ejected to heights of many thousands of miles through an atmosphere of probably greater and certainly equal density, but the shapes assumed by the hydrogen streaks are in- consistent with the idea that they can have been themselves ejected. For instance, the shreds of hydrogen observed by Prof. Young (some of which * Of Dartmouth, N.H., when the discovery was made. It has been hoped that Prof. Young, with the much more power- ful telescopeat Princeton, will make discoveries even exceeding in interest those which he made at Dartmouth, no OTHER SUNS THAN OURS. were thousands of miles long) were irregular in shape; had they really been travelling through a resisting atmosphere, at the enormous rate of 145 miles per second, they would certainly have been pear-shaped, rounded in front and tailed behind, like fire-balls in our own air. But they resembled, rather, the irregular streaks showing where our air has been rendered luminous by the passage of meteoric masses through it. Prof. Young's observation proved, in fact, that on that particular occasion the sun had shot out from his interior a flight of many thousands of bodies. The bodies themselves would not be visible, because the phenomenon was observed through a telespectro- scope, admitting only red light of the same tint as the red of glowing hydrogen. But the light from the heated hydrogen along the tracks of these ejected missiles would be clearly visible. The streaks would, of course, seem to ascend. For they would always be close up to the missiles producing them, so that their forward ends would advance, while their rear ends would seem also, to advance as the light gradually faded out along those parts of the track which were farthest from the missile. What Prof Young saw has been seen since, at various observatories. The sun, then, has the power of ejecting matter from his interior — presumably in volcanic explosions. Moreover, a calculation which I made respecting Prof. Young's explosion shows that the matter ejected on that occasion passed away from the sun with such velocity that it would never return to him. Those missiles were thenceforth akin to meteoric bodies travelling freely through space. We may fairly extend the evidence thus given respecting the one sun we are able to study to other suns — and the extension may be made to other suns in time as well as to other suns in space. If the one sun we are able to study, because he is comparatively SUNS AND METEORS. Ill near to us, and because he is a sun now, is able to eject flights of bodies from his interior to vast distances, and even to cast such bodies for ever away from him, the other suns which people space possess in all probability a similar power, and orbs which were suns in the remote past, possibly also orbs which will be suns hereafter, were or will be similarly active. Taking first the extension of the evidence given by the sun to bodies no longer suns, we see that what has already been suggested in other ways is confirmed by the evidence of the actual eruptive power possessed by the sun. We see that millions of years ago, when Jupiter and Saturn were active suns, they probably possessed the power of ejecting flights of bodies from their interior as the sun does now, and many millions of years ago, when our earth and her fellow terrestrial planets were sunlike bodies, they were similarly active (each in its degree). For it is, of course, obvious that though a body like Jupiter would have nothing like the sun's eruptive energy (in amount), such an orb would need nothing like that energy to eject matter from its interior never to return. So with a globe like our earth. The sun must eject a body with a velocity of 380 miles per second, that it may never return to him ; and Jupiter would have to impart a velocity of about forty miles per second to reject for ever a mass erupted from his interior ; but in the case of our earth a velocity of seven miles per second would suffice to carry ejected matter for ever away from her (apart, of course, from the chance of subsequent capture by accidental encounter with the parent orb, whose course the track of the ejected mass would always thereafter approach or intersect). Now, though no volcanic explosions which at present take place eject bodies from the earth with anything like this velocity, yet remembering the intense activity of an orb in the sunlike stage, as compared with the energies of the life-bearing stage, we see 112 OTHER SUNS THAN OURS, that even apart from the evidence given by solar explosions, and from the subsidiary evidence given by the meteoric paths, we might safely infer that the volcanic outbursts taking place during our earth's sunlike stage were probably quite sufficiently intense to eject matter for ever from her interior. If such an explosion as that of Krakatoa can take place now, outbursts of the mightier sort necessary for meteor-ejection may well have occurred when the earth was a small sun. We have similar actual evi- dence even in the case of the giant planets, for what- ever theory may be formed of the great red spot on Jupiter, there can be no doubt that a disturbance affecting an area nearly as large as the whole surface of the earth, and lasting seven years in full activity, implies most tremendous energies when Jupiter was in the sunlike stage of his career. As to the future, we cannot speak so confidently. We know not what the bodies are, if bodies there be, which will hereafter become suns. Possibly the great gaseous nebulae are forming into stars. It seems unreasonable, at any rate, to suppose that, as there are suns much younger than our own (Secchi's first order), as well as suns much older (Secchi's third and fourth orders), and suns long since dead (being dark), there are not also suns as yet unformed. Ranging through space we recognize in every star a sun, not only like our sun pouring out light and heat, but doing doubtless such other work as our sun is doing. If he pours out in a single explosion thousands of meteoric bodies, in the millions of years of his life, he must have poured out many millions of millions of such bodies. The millions of millions of other suns which people space must have done likewise. So that inconceivable numbers of bodies expelled from existent suns must now be traversing space. SUNS AND METEORS. II3 In such meteor streams — or comets — we find the explanation of those comets which reach our solar system from outside the planetary system. Some of the comets of long period maybe regarded as having had their origin from our own sun, but only those whose paths approach very near to his globe. For although planetary perturbations might prevent a body ejected by the sun from actually returning to him, as, if undisturbed, it must inevitably do, such perturbations could not possibly give to a sun- expelled body a path passing far from the sun's globe. The comets, therefore, or meteor systems which travel around the sun on orbits passing far outside the planetary system, and those whose orbits carry them away from our sun never to return, are ex- plained as flights of bodies ejected either from our sun himself (in the case of a very small proportion only) or from other suns. But among the suns there are some so much mightier than the rest that we might expect the meteor systems sprung from them to differ in marked degree from all others. I refer to the giant suns like Sirius, Vega, Altair, and others of Secchi's first order.* Sirius, judged by the quantity of light he emits, is probably at least a thousand times larger than the sun ; and we may infer that the other suns of the same order are in like degree superior to our sun both in size and in energy. Surely the meteor flights ejected from these giant suns would be as markedly distinct from those ejected by our sun and his fellows as these meteor flights are distinct from those ejected by the giant planets, and these in turn from those ejected by the earth and her fellow planets of the terrestrial order. In particular, the velocities of comets or meteor flights ejected from Sirius, Vega, and their fellows, * As classified by their spectra. I 114 OTHER SUNS THAN OURS. would be apt to exceed enormously the velocities belonging to meteors ejected from suns of the same order as our own. When I was first led to adopt the theory which I have here indicated I thought it likely some evi- dence might be obtained of meteor systems ejected from the giant suns. But no such evidence actually existed at that time — about twelve years since. Now, however, evidence of absolutely decisive nature, evidence not only confirming my theory, but explic- able — so far as I can see — in no other way, has been obtained. Five years ago Mr. Denning, of Bristol, announced that he had recognized some meteor systems which radiate for several months in succession from the same point in the star sphere, a result which seemed so surprising that at that time many rejected it. I rejected it myself for awhile. It seemed to me, in- deed, too good to be true. But it has since been shown to be undoubtedly true. Now, the same reason which forces us to regard the radiation of meteors during several hours from the same point, as proving that our earth's velocity of rotation is in- significant compared with the velocities of these meteors, compels us to regard the velocity of the earth's revolution as insignificant compared with the velocities of meteors which radiate during several months from the same region among the stars. In six hours the rotational motion of a point on the earth changes through a right angle ; in three months the motion of revolution of the earth her- self changes in direction in the same degree. But one motion has a rate of only a third of a mile per second even at the equator, the other has a rate more than fifty times greater. Mr. Denning's observation shows that there are meteor systems travelling many times faster than the earth in this swift rush round the sun. These meteor sys SUNS AND METEORS. IIJ tems can be no other than those which have been expelled from the giant suns. Hence finally we recognize, by direct evidence,* four orders of suns and four orders of meteors. First, earth-suns, long since dark, which expelled such meteor systems as those which have been recog- nized as earth-born. Secondly, giant planets, long since deprived of sun-like brilliancy, but not yet dark, which expelled such meteor systems as now travel on orbits passing near the paths of Jupiter, Saturn, Uranus and Nep- tune. Thirdly, bodies like our sun, which expelled and still expel such meteor systems as travel on orbits extending far beyond the solar system. Fourthly, bodies like the giant suns, which expelled meteor systems travelling with much greater velocities than could be imparted by our own sun oi his fellows of the same order. * I have said nothing here of the evidence given by the microscopic, chemical, and physical examination of meteors. Such evidence has, in reality, proved that those bodies were once in the interior of orbs in a sunlike state.' CHAPTER IX. COMETS AND METEORS. We find among men of science a singular mixture of caution and daring, degenerating sometimes into timidity on the one hand, and into rashness on the other. The scientific caution of a Newton, testing the theory of gravitation by line and measure, and calmly resigning it for awhile, because, as it chanced, line and measure were both inexact, may be com- pared with the noble daring of a H alley, boldly announcing that the comet of 1682 would return in 1758* on the strength of observations which, in our day, would certainly be thought insufficient to deter- mine a comet's period. The timidity with which the profound reasoning of Olmsted respecting meteors was rejected, till simple observations made that obvious which he had made certain, may be con- trasted with the rashness shown by those who have accepted the speculations of Laplace about the universe as though these were demonstrated theories. Comets, the most mysterious of all the bodies known to astronomers, have been subjects of most marked timidity and of most daring rashness of scientific reasoning. That men should have been * I am quite aware of the fact that the comet really returned in 1759, that is to say, that it was in 1759 that the comet passed its point of nearest approach to the sun. Halle/s prediction, however, named 1758, and made as it was when the theory of gravitation was in its babyhood, it was a very fair guess. 116 COMETS AND METEORS. II7 unwilling to formulate definite theories about these wild wanderers is; perhaps, natural enough. But the calm, uninquiring confidence with which ideas have been advanced and suggested respecting comets is not so easily explained. One of these ideas, regarded by many as if it were an established truth, I propose now to inquire into, — the idea, namely, that comets have been drawn from those paths on which they chanced originally to approach our solar system, by the perturbing influences of the giant planets, and have thus been, in certain instances, compelled to travel around the sun in elliptical paths, instead of the parabolic or hyperbolic orbits on which they had been travelling before. they were thus captured. I think I shall be able, first, to show that this theory is antecedently most unlikely ; then to prove that even if it had been the most natural and probable theory conceivable, it is entirely incon- sistent with observed facts, and, therefore, untenable. I shall then suggest a theory in its place which, were I to mention it just here, would probably be rejected at once as the wildest speculation imaginable. Possibly, introduced as it will be by a series of observed facts not otherwise explicable, it may not seem so repellent a little further on. But I shall ask the reader interested in matters cometic, not to turn to the end of this essay until he has read the beginning. We start from the conception that all comets originally entered our solar system from without. They come, say Heis, Schiaparelli, and others, who have advanced the Capture Theory, from out of interstellar space. Now, it is no valid objection to this view that it gives us no idea how cometary matter came to exist in interstellar space, for in all inquiries into the past condition of the celestial bodies we must always come short of their actual origin. Thus, in considering the past of our solar Xl8 OTHER SUNS THAN OURS. system we may start from a chaotic vaporous state, or from a past condition in the form of cosmical dust, or from a condition in which the vaporous and the dust-like forms are combined ; but if we are asked whence came the vapour or the cosmic dust we are obliged to admit that we cannot tell. If, hereafter, we should be able to say that it came from such and such changes in a quantity of various forms of matter, which we may represent by X, Y, and Z, we should still be unable to say how X, Y, and Z came into existence. So that I make no serious exception against the supposed origin of comets on the ground that it really leaves very much to be explained. Interstellar space is a convenient place to which to assign the origin of bodies so mysterious as comets. Cela explique beaucoup de choses. Almost anything might happen in regions of which we know so little, or rather, of which we know absolutely nothing. Yet it may be worth while to remark that, on the whole, the interstellar regions are less likely to be the regions whence comets originally came to visit suns and sun systems, than to be regions whither comets strayed after leaving originally the neighbour- hood of solar systems. The most probable idea about the interstellar spaces is that they are the most vacuous regions within the range of the sidereal system. The mere circumstance that comets came from out of them affords no better reason for regard- ing them as the original home of comets, than the circumstance that comets pass from the solar system into these interstellar spaces affords for rejecting that assumption. There is, in fact, simply no reason whatever for imagining that the place where comets came into existence is the vast unknown region around the solar system which we call interstellar space. Mast comets come to us from thence; as many comets are travelling into that unknown region COMETS AND METEORS. II9 as are coming out of it. To form an opinion about the origin of comets from no better evidence than their last journey (out of millions, very likely) can afford, would be as absurd as for a day-fly to reason that the river flowing past the home of his race came out of the sky because a few drops of rain came thence. Suppose, however, we admit that in interplanetary space there have been in th^ past, and still exist, such flights of meteoric matter as the theory we are con- sidering assumes. Let us grant them, also, such motion as may save them from what otherwise would inevitably be their fate, viz., a process of direct in- drawing toward the nearest sun, and consequent destruction (with mischief probably to his orb), after a period of time which must be regarded as utterly insignificant compared with the time-intervals mea- suring the duration of a solar system. It follows, then, that each flight of meteors would, in the long run, draw near some sun, without how- ever rushing directly upon him ; and, sweeping round his globe upon such path as chanced to result from the combination of its original movement and his attractive influence, would pass out again into interstellar space. This might happen tens, hun- dreds, thousands, or even millions of times, a comet either sweeping in a long elliptical orbit, with enor- mous periods of revolution, around one sun, or, if its velocity were slightly greater than that supposi- tion implies, rushing first round one sun, then out into the depths of space to visit another sun, then to yet another, and so on, flitting from sun to sun for ever, or until the kind of disturbance in which the holders of the theory we are considering believe, had changed this sort of motion into actual orbital circuit.* * I have here considered only two kinds of cometic orbit — the elliptic and the hyperbolic ; for a true parabolic orbit would I20 OTHER SUNS THAN OURS. In either case the minimum velocity with which a comet would be moving, when at any given- distance from our sun, would be determinable within a few yards per second. It is well known that the velocity with which a body travelling to the sun from an infinite distance (though one cannot, of course, con- ceive such a movement) would reach the sun, would not exceed by a foot per second the velocity with which a body would reach him after travelling from the distance of the nearest fixed star. So, also, the velocities of bodies moving in orbits reaching half as far from the sun as the distance of the nearest star, would be the same within a foot or so per second as the velocities with which bodies coming to the sun from infinity would reach the same dis- tance from him. If such bodies had originally a great inherent velocity, of course they would reach any given distance from the sun with much greater velocity. But this would not affect our estimate of the least velocity at that distance. Thus we know what the giant planets to which has been attributed the final capture of those comets which now form a part of the solar system, had to do. We can tell the precise velocity in miles per second, or, at least, the minimum velocity, with which our imagined meteoric flight would cross the orbit of Neptune, or Uranus, or Saturn, or Jupiter, as the case might be, before its capture. We know, in the case of each comet sup- posed to have been captured, the precise velocity of the comet at the distance of the planet which captured it— its special planet-master. The differ- ence is the amount of velocity which the capturing planet had to take away in order to effect the sup- posed capture. Observe that we are; here on sure ground, if the theory is sound. It is certain that a comet in coming be as unlikely, or rather as impossible, as a truly circular orbit among the planets. COMETS AND METEORS. 121 from remote interstellar space to the solar system would have at the distance, say, of Jupiter, a certain velocity. It is certain that a comet now travelling in a particular orbit, approaching at one point very near to the orbit of Jupiter, has at Jupiter's distance a certain velocity, very much smaller. Hence, it is certain that, if Jupiter captured that comet by dis- turbing it, as it approached him on the last of its many free visits to the sun, the giant planet must have deprived the comet of so many miles per second of its former velocity. All we have to do is to find out how the planet could do this ; in other words, how near the comet must have approached the planet to be thus effectively disturbed. These pages are not suited for the close and exact discussion of the case of any particular comet. I have elsewhere (in a paper which appeared in the " Proceedings of the Astronomical Society ") given the details for certain cases which have been re- garded as among the most satisfactory illustrations of the comet-capturing ways of the giant planets, and have shown that the theory is in those cases, and therefore in all, absolutely untenable, though so resolutely held. Still it may be well here to con- sider an illustrative general case — the simplest thai can be taken, and also the most effective, because the conditions are, in reality, much more favourable than they are in any known case. Imagine a flight of meteors to travel from inter- stellar space toward the sun until it reaches the dis- tance of Jupiter, and that when at that distance it chances to pass very close to the orbit of Jupiter, and at a time when Jupiter himself is very near the place where the meteor flight crosses his track. Observe that the chances against each one of these contingencies are enormous. If we conceive a sphere around the sun, girdled by Jupiter's orbit, the meteor flight in its course sunwards might traverse the sur- 122 OTHER SUNS THAN OURS. face of that sphere (or, which is the same thing, might traverse the part of its course where it is at the same distance as Jupiter from the sun) anywhere, and we are supposing that it traverses that surface close to a particular girdling circle (technically, a " great circle " of the sphere). Suppose that by " close " we mean within a million miles ; then the imaginary girdle of the sphere through which the meteor flight must pass to fulfil the required condi- tions is two millions of miles broad. The sphere itself has a diameter of some 960 millions ' of miles, and by a well-known property of the sphere,* its surface is 480 times greater than that of the girdling strip. The chance is but one in 480 that any meteor flight coming from interstellar space toward the sun will be within a million miles of Jupiter's orbit when at Jupiter's distance from the sun. Then Jupiter's path has a circuit of more than 3,000 millions of miles. Thus the chance that at the moment of the meteor flight's passing the orbit, Jupiter will be within a million miles on either side of the place of passage, is as two in 3,000, or one in 1,500. But the chances that both these relations hold is only as one in 1,500 multiplied by 480, or as one in more than 700,000. Thus, assuming — though the case is other- wise — that a million miles would be an approach near enough for capture, still only one meteor flight out of 700,000 which came from outer space could be captured by Jupiter. This, however, is but the mere beginning. We may admit that millions of times as many comets or meteor flights approach our system as the planets have captured ; and if so, we need recognize no special force in any such considerations as have just * The property is this : that the surface of a sphere exceeds the surface of a girdling strip, such as we are considering, in the same degree (if the strip is relatively narrow) that the diameter of the sphere exceeds the breadth of the strip. COMETS AND METEORS. 123 been presented. I only advanced them to suggest the conditions which are, as it were, essential for the process of comet-capturing by a giant planet. Arrived at Jupiter's distance from the sun, the meteor flight from interstellar space will have a velo- city of about eleven miles per second. Now let us inquire what its velocity must be reduced to in order that it may thenceforth be compelled to travel in a circle around the sun. As a matter of fact, all the members of Jupiter's comet-family travel in orbits whose remotest parts are near Jupiter's orbit, and to give a comet such an orbit as one of these, much more must be done in the way of reducing velocity than is necessary merely to make the meteor flight from outer space travel thenceforth in a circle at Jupiter's mean distance. We are taking, in fact, a very unfavourable case for our argument. Still, the velocity must be reduced, even in this case, by nearly three-tenths, or by more than three miles per second. Now Jupiter's power to withdraw velocity from a body in his neighbourhood is measured by his power to impart velocity. In fact, both processes are but different forms of the same kind of work. Precisely as we say that the sun can communicate a velocity of 382 miles per second to a body approaching him from interstellar distances, and that therefore the sun can withdraw such velocity from a body leaving his sur- face at that rate and eventually bring such a body to rest out yonder in interstellar space, so can we make a corresponding statement for any planet, — Jupiter or Saturn, the earth, our moon, and even for the least of all the asteroidal family (supposing only the mass and size known). In the case of Jupiter, for instance, we find that the utmost velocity he can impart to a body reaching him from external space is about thirty-six miles per second. That, at least, is the velocity with which such a body would reach 124 OTHER SUNS THAN OURS. the visible surface of the planet. What the velocity might be with which the real surface, far down below the visible envelope of clouds, would be reached, we do not know, — not knowing where that surface lies. In the case of our own earth, the velocity with which a body would reach the surface, if brought thither solely by the earth's action from interstellar space, would be a little over seven miles per second, or more than twenty-seven times greater than the velocity of the swiftest cannon-ball. But while Jupiter — to keep for the moment to our giant planet — has thus, theoretically, the power of giving or taking away a velocity of thirty-six miles per second, he is not practically able to do anything of the sort. He is not left to draw matter to him- self, or to act on the recession of matter from him- self alone. The bodies which come near to him from outer space have been drawn by solar might within that distance from the sun, and almost the whole velocity they there possess is sun-imparted. We have seen that it is some eleven miles per second. Now, manifestly, this greatly affects Jupiter's power of imparting or withdrawing velocity. Both pro- cesses require time, and it is clearly impossible for Jupiter to produce anything like the same effect on a body rushing past him with a sun-imparted velocity of eleven miles per second as he would produce on a body left undisturbed to his own attraction. Jupiter's action at any moment is the same, whether the body is moving or at rest ; but the number of moments is very much reduced owing to the swift rush of the body past the planet. To use the old- fashioned expression of the first students of gravita- tion (an expression which has always seemed to me amusingly quaint), the solicitations of Jupiter's attrac- tive force are as urgent on a swiftly rushing body as on one at rest ; but if a body will not stay to hearken to them much less effect must be produced. COMETS AND METEORS. I2S In all this part of my reasoning, I may remark, I am not pleading a cause, but indicating what every student of celestial dynamics knows. We may fairly regard twenty-five miles per second as the utmost velocity that Jupiter can impart or take from any body coming out of interplanetary space past him, as close as such a body can pass without being actually captured. Moreover, in every possible case, Jupiter can only abstract or add a small portion of this amount ; for this reason, simply, that in every possible case there will be first an action of one kind (abstraction or addition of velocity), and afterward an action of the opposite kind (addition or abstraction respectively). It will be but the difference between these effects, in most cases very nearly equal, which will actually tell on the body's future period of revolution around the sun.* This makes an enormous reduction on Jupiter's potency to modify cometic revolution. Certainly ten miles per second is a very full estimate of the velocity he can abstract or add in the case of a body passing quite close to his apparent surface. But even this may seem ample. Seeing that a loss of three miles or so per second would cause a body which had reached Jupiter's distance from the sun, after a journey from out of interplanetary space, to travel in the same period around the sun as Jupiter himself, and since we seem to recognize a power in Jupiter to abstract ten miles per second, it would seern as though Jupiter's capturing power were in fact demonstrated. But while, to begin with, the close approach re- quired for this capturing power to exist is something very different from that approach within a million * As distinguished from the orbit. The orbit might be largely affected even in a case where the velocity at Jupiter's distance remained absolutely unchanged ; but in this case the period of revolution would remain the same. 126 OTHER SUNS THAN OURS. miles which I before considered, there is a much more important difficulty to be considered, in the circumstance that we have thus far dealt with Jupiter's capturing power on one body, not on a flight of bodies, such as a comet approaching from interstellar space is held to be, according to the theory I am discussing. Let us take the former point, though the least important, first. At Jupiter's apparent surface the actual maximum velocity which the planet could give to a body approaching from a practically infinite distance would be about thirty-six miles per second, and we reduced the actual maximum effect on a body passing Jupiter very closely, under such conditions as actually prevail in the solar system, to ten miles per second. Let us see what would be the corresponding numbers in the case of a body passing within a million miles of him, remembering that even that would carry such a body right through Jupiter's system of satellites, the span of that system being about four and a half millions of miles. Since a distance of one million miles exceeds the distance of Jupiter's surface from his centre nearly twenty-five times, it follows (I need not explain why; mathematicians will know, and for non-mathematicians the explanations would be tedious and difficult) that the velocities which Jupiter can give or abstract at the greater distance would all be reduced to little more than one-fifth those deter- mined for Jupiter's surface. So, instead of ten miles per second, we should get but two miles per second, as the greatest Jupiter could abstract from a body approaching him within a million miles. And this would not be sufficient reduction to make such a body travel thenceforth in Jupiter's period, still less in one of the much shorter periods observed through- out what has been called Jupiter's comet-family. But the other difficulty is altogether more serious. COMETS AND METEORS. 127 A comet approaches Jupiter, on the theory we are dealing with — and, indeed, the same may be assumed on any theory — as a flight of scattered bodies. Either this flight is so close as to be in effect, be- cause of mutual attractions, a single body, or it is not. If it is, the flight will not be broken up by Jupiter's action ; and, if not so broken up, will re- main for ever after a united family. But if, as is more in accordance with observed facts, the cometic flight is so large that the attraction of the flight, as a whole, on the separate members, can be overcome by Jupiter's action, then not only will the flight be broken up, but the orbits given to different members of it by Jupiter's disturbing action will be widely different. Suppose, for example, the extent of the flight to be such that the parts coming nearest to Jupiter approach his centre within fifty thousand miles (a very close approach indeed to his surface), while those parts which are remotest from him at the time when the flight, as a whole, is nearest, came only within sixty thousand miles from his centre. Then, in round figures, the reduction of velocity of the nearer members of the flight will be greater than the reduction for the farther members, a^six exceeds five. Supposing, for argument's sake, the former reduction to be three miles per second, as it must be to make those members of the flight travel thence- forth in Jupiter's period round the sun, then the reduction for the outermost members would be but three and a half miles per second ; or thence- forth one set of meteors formerly belonging to the comet would have at Jupiter's distance a velocity of eight miles per second (eleven less three), while another set would have a velocity of eight and a half miles per second (eleven less two and a half) at that distance. This means that thenceforth the mean distance of the latter set from the sun would exceed the mean distance of the former set about as nine 128 OTHER SUNS THAN OURS. exceeds eight.* Since the former set would thence- forth be travelling at Jupiter's distance, or about S"2 times the earth's, the latter set would be travelling at a mean distance greater by one-eighth of this, or "65 of the earth's distance, say some sixty millions of miles. The latter set would be at their nearest to the sun when at Jupiter's distance, would pass sixty millions of miles farther away to their mean distance, and as much farther away still at their greatest dis- tance. Practically, then, even in this case, as favour- able for capture as can be well imagined, the capture, though effected, would result in spreading out the comet, which had arrived as a compact flight of meteors ten thousand miles only in span, over a region one hundred and twenty millions of miles broad. It is hardly necessary to say that nothing like this is observed in the case of any member of Jupiter's comet-family. We know that along their track meteors are strewn to distances which, in some cases, may well exceed even the enormous distance just named ; but they lie along the track, not ranging more than a few hundred thousand- miles on either side from the path of the comet's head. This means that the orbit of every single meteor of such a system has practically the same mean distance from the sun. The difficulty last considered is simply fatal to the theory that the comets forming what have been called the comet>-families of the giant planets were captured by those orbs in the way imagined by Heis, Schiaparelli, and others. We must seek for a different explanation, if we are to account for the peculiar relations of these comet-families at all. It * The simple law is, that for two bodies having different velocities at the same distance from the sun. the mean dis- tances from him differ as the square of those velocities. Now, the square of eight and a half is seventy-two and a quarter ; that of eight is sixty-four. COMETS AND METEORS. 129 may be that the pecuHarity, Hke many others pre- sented by comets, may not admit of being explained. The considerations I am about to advance may to many appear not altogether convincing ; neverthe- less, as they involve the study and discussion of known facts, they are worth investigating, quite apart 'from all questions of the validity of the theory with which I associate them. Observing that the giant planets have each their comet-family, we may safely infer that the sun also has his special family of comets ; that is, a family the dominion of which he does not in any sense share with the giant families. The comets which we should thus regard as specially solar are those whose paths approach exceptionally near to his globe. Among numbers of comets which come from out of interstellar space toward the sun, and, sweeping around him, pass away again into the depths from which they came, many have paths passing so far from his globe that we cannot regard them as in any special way associated with him. Bodies coming casually, so to speak, from outside regions would have just such paths. So that of many comets, not belonging to the comet-families of the giant planets, we may say that neither do they belong to the comet-family of the sun. Yet even these teach something. Whatever theory we adopt as to the origin of comets, it must give an account of these comets, as well as of those which, passing very near to the globe of the sun, may be regarded as belonging specially to him, and those others which we assign as the special dependents of the giant planets. Now, taking the two last-named classes, we re- cognize in the movements of the members of each class evidence of the introduction of these comets into the solar system, through the intervention, in some way, (i) of the giant planets in the case of one K 130 OTHER SUNS THAN OURS. class, and (2) of the sun in the case of the other class. We have seen that the giant planets could not have introduced their comet-families from out of interstellar space by perturbing influences. We may infer with almost equal probability, or almost with certainty, that neither did the sun introduce his comet-family by drawing them from out of interstellar space. Since, then, the sun and the giant planets did not introduce their special comet-families from inter- stellar space, yet did most manifestly introduce them in some way, where else can these comets have come from but from within the orbs of the sun and of the giant planets respectively ? At first sight this theory seems so strange and fanciful that we are almost deterred from examining it further by its apparent grotesqueness. We seek about for a way of escape from so wild a theory. We look back to a remote period when, in accord- ance with the ideas of Laplace, the sun's mass ex- tended far beyond the present orb of the sun, and the giant planets also had orbs extending even as far as the orbits of their outermost satellites. Un- doubtedly, if a flight of meteors in that far distant period rushed through the outer vaporous surround- ings either of sun or of giant planets, the effects imagined by Schiaparelli and by Heis might have been produced. The diminution of the velocities of the meteors forming such a flight might well be far more effective than in the case we have hitherto considered of free space around a planet's globe. But we may regard this theory respecting- the introduction of comets into the solar system as one which may wait its turn until the other, of ejection, strange and fanciful though it may seem, has been examined. For there is nothing in the capture theory, considered in itself, to invite us specially to its adoption. It gives no account whatever of the COMETS AND METEORS. I3I actual origin of comets. It only suggests how, having somehow come into existence in interstellar space, comets would be drawn sunward, and might be captured by the sun or by planets. If to this inherent difficulty in Schiaparelli's theory we are to add all the difficulties involved in the supposition that the sun and the giant planets were once much larger than they now are, and that being thus large they were able to capture comets by actual interrup- tion of their movements, we may at least consider that before discussing such views, before attempting to carry back our thoughts over the practically inter- minable time-intervals involved in such a process, it may be well to examine a theory which, though startling at a first view, promises to explain some- thing more, if confirmed, than the scarcely less startling theory of comet-capture by expanded sun and by expanded planets. Suppose that instead of looking into remote regions of space, and toward far-off periods of time, we examine meteoric masses, and inquire of them whence they came. We cannot expect each me- teorite to have a story to tell ; but after a goodly number have been examined, we may light upon one speaking with tolerable clearness respecting its origin. Our first studies shall be with the micro- scope. Now, passing over a number of microscopic studies of meteorites which are suggestive enough, but not decisive, we come on the strange fact that certain meteorities show under the microscope the clearest evidence of having once been in the form of tiny globules of molten metal, numbers of which have become agglomerated together. The eminent mi- croscopist and mineralogist Sorby, of Sheffield (England), asks respecting these particular meteors, where else could they possibly have existed in the form of metallic globules (liquid) except in the K 2 132 OTHER SUNS THAN OURS. interior of a body like the sun ? In the interstellar spaces intense cold prevails. In rushing close past the sun a meteoric mass might be molten, but would scarcely be vaporized, even though the orbit of the flight passed very near the sun's surface. But the meteorites which have visited our earth have not been associated with comets passing near to the sun. Manifestly the chances are very small that any meteorite following in the train of a comet like Newton's or the comet of 1843 — that is, a comet travelling close past the sun — would ever reach the earth. But Sorby found microscopic evidence such as I have described in quite a large number of meteorites which he examined. At any rate, the assumption for the moment, that such meteorites had their origin within the interior of a body like our sun, accords well with the theory we have had suggested to us, that comets and meteor flights (kindred bodies) came from within the orbs with which we still find them associated. Turn now to the chemical analysis of meteorites. Here the evidence is perhaps even more suggestive. Masses of meteoric iron being placed under the air- pump, hydrogen which had been present in their substance — occluded in the iron, as it is technically expressed — has come out in such quantities that Professor Graham (of London) considers the amount fully six times as great as could be occluded in the substance of iron by any process known to chemists or physicists. This Lenarto meteor, he says, has brought to us across the interstellar spaces the hydrogen of the fixed stars. In other words. Pro- fessor Graham could see no other interpretation of the presence of so much hydrogen within the sub- stance of this mass of meteoric iron than that the hydrogen had been forced into the iron while yet within the interior of a star. We know that beneath the visible surface of our sun there must be both the COMETS AND METEORS. 133 vapour of iron and hydrogen at enormous pressure. Under such conditions alone could masses such as the Lenarto meteorite be formed. Professor Graham, therefore, assumed confidently that the Lenarto meteorite and others of the same sort were formed in the interior of a body like our sun. He rejected, rightly, the idea that it was in our sun himself that the. meteorites of that class were formed. For the chance of any meteorite ejected from the sun reach- ing our earth is but about as one in twenty-two hundred miUions. The greater number' of the sun- ejected meteorites must, he saw, have been ejected from the interior of the other suns which people space. There are hundreds of millions of such suns even within the range of telescopic vision ; millions of millions doubtless exist ; so that if we once admit the possibility of the ejection of meteoric masses from within a sun or star, we recognize the probability, or rather the certainty, that there must be billions of billions of such masses travelling amid the interstellar spaces. All this was reasoned out thus before it had been shown that suns ever do eject masses with sufficient energy to carry them beyond the attractive influences of their parent orbs ; nay, Sorby and Graham expressed their views respecting the origin of some meteorites when it seemed utterly unlikely that we ever should get the evidence of stellar erup- tive powers which that theory requires. But such evidence has now been obtained. Pro- fessor Young, of Princeton, N.J. (then of Dartmouth, N.H.), was the first, in 1872, to obtain evidence of the actual ejection of matter from the sun's interior with velocities sufficing to carry such matter for ever away from him ; but the evidence was decisive, and since then kindred observations have been frequently made. What Young saw, indeed, was apparently the ascent of filaments of hydrogen, at an average 134 OTHER SUNS THAN OURS. rate of nearly two hundred miles per second ; but it was easy to see that the irregular streaks of hydro- gen were not themselves the ejected matter. If a thin gas like hydrogen could rush through the region immediately above the sun's visible surface at the rate of two hundred miles an hour, — which I reject as incredible, — the shape of such hydrogen missiles would be such as to indicate very clearly the resist- ance they were encountering. They would be pear- shaped, the ^rounded part of the pear in front, like fireballs in our air. But these were irregular streaks, like the luminous tracks of meteors, and such doubt- less they were. A flight of masses of considerable density must have been shot out on that occasion, and on other occasions when similar phenomena have been observed, and rushing through the hydrogen in the sun's neighbourhood, caused the gas to glow along their track, just as fireballs in our air leave behind them long luminous trails. The rate at which these missiles advanced could be inferred from the rate at which the luminous trails followed them. Calculation, in which the sun's retarding action was taken duly into account, showed that the matter thus expelled from the sun left his surface at a rate of not less, probably, than five hundred miles per second. The ejected matter left the sun, then, never to return, and in the form of precisely such a flight of meteoric missiles as microscopic and chemical researches had shown to be travelling through the interstellar spaces. When we consider the three lines of evidence, and note how independent they are of each other, we see that the theory of the ejection of masses akin to meteors from the suns which people space is rendered all but certain, independently of any line of d priori reasoning which had led us to look for evidence of such processes. Certain meteors have shown under microscopic study that they were certainly once in a COMETS AND METEORS. 1 35 condition such as could hardly exist except in the interior of a body like the sun ; others have shown under chemical analysis that they must have been ejected from the interior of a sun ; and now we have evidence showing that from our sun, and therefore presumably from his fellow-suns, the stars, flights of missiles akin to meteoric bodies are ejected from time to time with velocities sufficient to carry them into interstellar space. It seems reasonable to infer that here we have the solution of our difficulty ; we see that the sun, at any rate, has power to eject at times, from his interior, flights of meteoric masses, such as we recognize in the streams of meteors which exist within the solar system, and that the velocity of outrush is in some cases so enormous that the masses thus ejected can never return to the sun, but pass away through interstellar space. We find also that meteoric streams, which we are thus led to associate with the solar eruptions, are also associated with comets, every known meteoric stream travelling, probably (as many certainly do), in the track of a comet. Now, knowing the small masses of many comets, it is no very wild thought to suggest that those comets whose present orbits carry them close to the sun were originally expelled from his own interior. Assuredly the flights of missiles which we know to be at times driven from his interior are in all respects akin to what we know many comets actually to be, akin in structure, akin in mass, and akin probably in condition. For in whatever re- spects the coma and tail of a comet may seem unlike mere meteoric masses, we know that such peculiarities of condition are due to solar action, and that a flight of meteoric masses ejected from the sun himself would as certainly present these peculiarities under subsequent solar influences as any other flight of meteoric masses not ejected originally from the sun. 136 OTHER SUNS THAN OURS. May not this reasoning be extended to the giant planets, either in their present demonstrably some- what sunlike state, or in those past stages of their career when they were veritable suns, though small ones ? In the great red spot of Jupiter, however, we have had evidence of even a present intensity of eruptive action by which meteoric and cometic matter might well have been ejected in such sort as to pass for ever beyond the control of the giant planet. At any rate, the great disturbance suggests, by parity of reasoning, that within comparatively recent times Jupiter and Saturn have possessed the necessary expulsive power. It must be remembered that thus to eject matter with velocities sufficient to carry it for ever away, Jupiter and Saturn would not need anything like the same ejecting power which the sun has to exert to expel matter for ever from within his globe. They are much weaker than the sun, but for that very reason they would need to exert much less eruptive force, seeing that it is their own attractive power they have to overcome, and that that is weaker in even a greater degree than probably is their eruptive power. Now, there is a family of comets attending in a sense on Jupiter, and another family attending similarly on Saturn, precisely as we should expect them to do if originally expelled from the interior of these planets. After such expulsion, though free to pass away for ever from their parent planets, they would not be free to pass away for ever from the solar system. They would be thenceforth attendant on the sun, but with this peculiarity, that no matter what perturbations they underwent, their paths would always pass near to the path of their parent planet. Even if in some future circuit a comet of this sort came quite close — as it very well might — to the planet it originally started from, it would still, though very much disturbed, follow a path possessing this COMETS AND METEORS. 1 3/ characteristic, however different from the path which it had before traversed. After many millions of years, indeed, it might happen, perchance, that resistance encountered in its movement around the sun, however ineffective to affect its orbit appreciably in a few thousands of years, would reduce the span of its circuit. But even then it would still be pos- sible to classify a comet whose orbit had been so changed, with the family of comets to which it had originally belonged. Now we find that among the periodic comets attending on the sun nearly all belong to families which have long since been relegated to the giant planets. There is a family of comets every member of which has an orbit passing very near to the orbit of Jupiter ; another family every member of which can be similarly associated with Saturn ; others depending in the same way on Uranus ; others on Neptune ; and, in fact, so fully has this sort of relation been recognized, that the idea has been thrown out that a planet travelling outside the orbit of Neptune, but as yet unknown, might be detected by the movements of a comet intersecting the great plane of planetary movement far beyond Neptune's orbit. It may be mentioned, indeed, in passing, that the comet of 1862, which has been associated with the meteors of August 10 and 11, intersects the plane of planetary movements at a place about as far beyond the orbit of Neptune as that orbit is beyond that of Uranus ; and that it has been held probable that at that distance a giant planet as yet undiscovered may travel. The existence of the comet-families of the giant planets can scarcely be explained v/ithout assuming that which we have thus been led on another line to recognize as probable, — the ejection from the giant planets of masses of matter, in eruptions akin to those taking place in the sun. Whether such 138 OTHER SUNS THAN OURS. eruptions take place now in the giant planets, or not, would be difficult to prove ; for although we have evidence of tremendous disturbances, we have nothing to show conclusively that these would suffice to eject matter for ever from within these planets' globes. Whether a careful study of the region out- side the discs of Jupiter and Saturn (the planets themselves being hidden by opaque discs) would decide the point, I am not prepared to say ; but I am certain that the edges of the discs of the giant planets are worth much more careful study than they have yet received. , But undoubtedly most of the comets of Jupiter's family must have been added to the solar cometic system hundreds of . thousands if not millions of years ago. Quite possibly both Jupiter and Saturn still eject matter from time to time with such velo- cities from their interiors that it passes away never to return to them. In this, as, in many other features, Jupiter and Saturn are still somewhat sunlike. But they have passed their truly sunlike youth. They tell us what our own earth was like when she was young. We may trace back her history, however, even to the sunlike state. The same law which we applied to the giant planets may be applied also to her. Her eruptive energies must have been very much less active, even in her sunlike youth, than those of the sun now ; but the force against which she had to work (her own attractive energy) was much less potent too : nay, it may probably have been less potent in even greater degree. Just as the moon in her volcanic youth upheaved her surface much more than the earth upheaved hers, because, though the moon was weaker, her subterranean energies had so much smaller downward-tending action of gravity to contend against, so it may well be that the smaller a planet when in its sunlike state, the more easily did eruptive forces eject matter COMETS AND METEORS. 139 beyond the range of the planet's attractive forces. In this case every planet at that stage of its career, as well as every sun, gave birth to cometic and meteoric systems, each after its own kind : solar comets being large ones like those which astronomers have not been able to associate with the planets' comet-families ; the comets ejected by the giant planets coming next in order of size ; and the comets ejected by smaller orbs, like the terrestrial planets, moons, asteroids, and so forth, being pro- bably too small to be discerned even with telescopic aid. CHAPTER X. WHENCE CAME THE COMETS? Although the astronomer has achieved many successes in studying comets, yet these objects still remain outside the surveyed fields of astronomy — now, as in the old days when men spoke of sun and moon, planet and stars, as including all the members of the heavenly host. The two comets now shining in our skies [l886] illustrate the present position of cometic astronomy. They have appeared without warning, we know not whence ; they have not until now been known to astronomers as travelling on recognized orbits and in definite periods ; and even hereafter, though the astronomer may determine their orbital motions and calculate the time when either should return, he cannot be sure that they will not be dissipated into unrecognizable portions before that time arrives. I do not propose to remark here upon the probable nature of comets, or upon the possible interpretation of the various phenomena they present. The only circumstance in regard to them which I shall take into account in what follows is that close relationship between comets and meteor-streams which was esta- blished in 1866 by the combined labours of Schia- parelli, Adams, and Tempel. I shall treat this kinship between comets and meteors as rendering certain or highly probable the four following pro- positions : — WHENCE CAME THE COMETS ? I4I (1) Every meteoric stream follows in the train of some comet, large or small, which either exists now or has been dissipated, as Biela's comet was, leaving only its meteoric trail to show where it once travelled. (2) Every comet is followed or preceded by a train of meteors (this train has nothing to do with the comet's tail), extending over a greater or less portion of the comet's orbit, according to the length of time during which the comet has existed. (3) All meteoric bodies, from those which exist as the finest dust to the largest meteorites, hundreds of pounds in weight, may be regarded as bodies of the same kind, differing from each other, indeed, in con- stitution as they obviously do in mass, just as planets and asteroids dio, but all to be interpreted — if they can be interpreted at all — in the same general way. We may in some degree illustrate the nature of the assumptions here made in the three following assumptions which an insect, who had observed the phenomena of rain, cloud, mist, snow, &c., might be supposed to make : (i) Every shower of rain im- plies the existence of a cloud ; (2) every cloud implies the descent, at some time or other, of rain, greater or less in quantity and heaviness ; and (3) all drops of water, from the tiniest water vesicles in a cloud to the heaviest rain- drops, are of the same kind, differ- ing oply in shape or in size ; snowflakes also, as formed of water particles in a changed form, must be put in the same class. And as the insect, by studying the relations which exist between clouds and rain, might be led to form an opinion whence clouds come, which would tell him also (as we know) whence rain comes,* so per- * To us, who know how clouds and rain are really produced, this imagined inquiry of the insect may seem trivial. But man had advanced far in scientific research before he had learned anything about the source and nature of rain, hail, snow, cloud, 143 OTHER SUNS THAN OURS. haps may we, by studying the relations which exist between meteor-streams and comets, be led to form an opinion whence comets (which are meteor collec- tions) originally came. The very first suggestion ever made respecting the origin of comets came, indeed, from such considera- tions as I have mentioned above. Schiaparelli, to whom we owe the happy guess, and the beginning of its confirmation as a useful truth, that meteors are bodies following in the tracks of comets, threw out the idea that comets, regarded as flights of meteors, may be travelling in multitudes through the inter- stellar depths, and be from time to time drawn out thence by the attraction of our sun. He pictured our sun, in his swift rush onward with his train of planetary attendants, as coming .into ever-fresh regions of comet-strewn space. A comet or meteor flight drawn towards him by the sun would approach the solar system on a path which may be described as casual. It might cross the general plane near which all the planets travel, at any point, the chance that that point would lie near a planetary orbit being very small indeed. Supposing the point where the meteor flight crossed that important plane — the life- plane of the solar system — to be on or near a planetary orbit, the chance would still be very small that the meteor flight would cross there at a time when the planet to which that orbit belonged was near that particular point. The chances would, in fact, be millions of millions, or rather of billions, to one that the meteor flight would visit our solar system without coming near any planetary body, in which mist, and fog. The whole subj.ect was as completely mys- terious, for example, to all the writers whose works were in- cluded by the Jews among their sacred books (in probably all their ancient documents), as were the phenomena of comets, which with them were veritable angels or messengers from Yahveh. WHENCE CAME THE COMETS ? I43 case it would pass out from our solar system again, never to return to it.* But, if a meteor flight did chance to come very close indeed to a planet of ade- quate mass, the flight might — said Schiaparelli — be captured. The planet might abstract so much of the comet's velocity as to leave only a balance corre- sponding to motion in a closed or elliptic path ; and on such a path would the meteor flight or comet necessarily travel thereafter — unless, perhaps, after many revolutions of each, the planet at some subse- quent encounter undid the work which it had accom- plished when first it approached the comet. So far Schiaparelli reasoned soundly on the basis of his assumption. I~ say assumption of set purpose ; for it is altogether a mistake to regard the idea thus thrown out by Schiaparelli as if it were a theory. His idea that meteors follow in the track of comets developed into a theory when it had been tested and confirmed by observation, and may now be regarded as a demonstrated and accepted theory. But the case is altogether different with the idea, or rather the mere fancy, that meteor flights are travelling hither and thither through the star-depths like fish in the depths of the ocean. However, his reasoning was thus far correct, assuming the meteor flights to exist and move within the interplanetary depths as he imagined. But beyond this Schiaparelli did not reason quite correctly. A single meteoric mass, or even a small meteor flight, might be introduced into our solar system in the way suggested by Schiaparelli ; for undoubtedly the giant planets possess the power he * Never J because, by the nature of its supposed indrawing, it possessed relative motion of its own before it began to be drawn in, and the sun could not take from it that relative motion. He would impart motion, and take such imparted motion away again, leaving untouched the original motion. Another way of putting this is to say that the comet's path would be hyperbolical. 144 OTHER SUNS THAN OURS. attributed to them, and if a body from without came near enough to any one of them, could so reduce its velocity as to change its path from the hyperbolic (or unclosed) form to an elliptic or closed orbit. And thenceforth such a body would travel around the sun systematically, on a more or less eccentric path passing very near the orbit of the planet by whose influence it had been originally introduced into the system. But a giant planet could do no more. It could not generate a meteor-stream in the way suggested by Schiaparelli. So soon as we test the matter by mathematical analysis, we find that approach so close would have to be made to a planet that a single body might be forced into a closed path, and it is certain that a flight of bodies large enough to pro- duce any of the known meteor- streams would have its components very widely scattered by the planet's perturbing action, simply because the different com- ponents of the flight would be exposed to very differ- ent degrees of disturbing action. This I have shown mathematically, and my de- monstration has not been questioned — though Pro- fessor Young, of Princeton, N.J., in admitting the validity of my reasoning, suggests the possibility that some way may hereafter be found for eluding the difficulty. But then Professor Young holds the strange idea that Schiaparelli's speculation as to the origin of comets and meteor- streams is an accepted theory ; and, labouring under this delusion, imagines that there must be some way of meeting objections to it, even though they may be mathematically demonstrable. But it is worthy of notice that Schiaparelli's fancy, even if accepted, would prove nothing about the origin of comets and meteors. To say that they came from out the interstellar depths on hyperbolic paths is to assert what can be disproved by mathe- WHENCE CAME THE COMETS ? I4S matical demonstration. But, if it could be proved what would it amount to ? Merely to this — that comets which now travel on closed paths once trav- elled on endless paths. We are no whit nearer the explanation of their origin. If the interstellar depths are crowded with meteor flights, we have to ask whence the meteor flights came. To say that fish which have been drawn from the sea were originally swimming about in the sea is surely not to add much to our knowledge about fish. It may be urged, however, that comets and meteor- streams are simply the material left unused after the various solar systems in our galaxy had been formed by processes of meteoric aggregation. Unfortunately for this explanation, the comets and meteor systems we have to explain are precisely those which, had they existed from the earlier ages when our solar system and its fellows were forming, would have been the first to be gathered up. For they are those which pass near the orbits of various planets, some near the orbit of Jupiter, some near that of Saturn, or of Uranus, or of Neptune, and about 400 which pass near the orbit of our earth. These comets, with their associated meteor systems, would have had less chance of escape than any others, during the millions of years belonging to the formative process of our solar system. Yet those are precisely the comets and meteor systems which we chiefly need to interpret. Suppose that, instead of making mere guesses, we consider the actual facts, and open our eyes to the views which seem to be suggested by such facts. I take first the millions of meteors encountered by the earth each year, and the hundreds of earth-cross- ing meteor systems already recognized. Taking for our guide proposition (i), we are led to the conclu- sion that in remote ages there were hundreds, if not thousands, of comets whose tracks crossed the track 146 OTHER SUNS THAN OURS. of the earth, or at any rate approached very near to it. That some of these comets thus crossed the earth's track casually, that is through mere chance coincidence, we may well believe. Nay, this is known, as will presently be seen. But if all did, then must there have been millions of millions of comets in remote times, to account for so many chancing to cross the earth's track — with this start- ling circumstance to be considered in addition ; that ninety-nine out of a hundred of those whose paths did not cross the earth's track have entirely disap- peared, while a considerable proportion of those which do cross that track (and which, therefore, have been exposed for millions of years to an extra risk of destruction) remain. This idea we may safely reject. But, if we do, then we have to account for a special earth-crossing family of comets and meteor-streams, without going outside to look for the origin of such bodies — for the moment we go outside we encounter the difficulty which has just driven us from any merely casual interpretation. In other words, we must look to the earth herself to explain the great majority of these earth-crossing systems. In this way Meunier and Tschermak were driven to look to the earth herself for the origin of meteorites. Proposition (3) above enables us to apply their reasoning, specially directed to particular classes of aerolites, to all classes of such bodies, to all meteors, down even to the tiniest falling stone, only visible perhaps in the field of a powerful telescope. Not all these bodies, but a goodly proportion, must have been generated in some specially terrene manner. We have actually no possible way of explaining the terrestrial origin of any meteors but in volcanic outbursts. Moreover, we are obliged to set the time when such outbursts took place very far back in the WHENCE CAME THE COMETS ? I47 past, seeing that at present the volcanic forces of the earth, even as manifested at Krakatoa recently, pos- sess nothing like the power necessary for the ejection of matter beyond the range of the earth's back- drawing power. Looking,' however, at the immense extrusive power of the volcanoes of the tertiary era, when basaltic lava covering hundreds of thousands of square miles to a depth of ,1,000 to 1.4,000 feet were poured forth, we can conceive the still mightier energies of volcanoes in the secondary era, their still more tremendous power in the primary era, and so, passing backwards to millions of years beyond the first beginnings of life on the earth, we can even picture to ourselves volcanoes ejecting matter with velocities of ten or twelve miles per second. With such velocities flights of ejected particles would pass beyond the earth's attraction, and if she were the only body in the universe, such ejected matter would travel away from her never to return. But, although such expelled bodies would never re- turn to the earth, they would not escape from the solar system. To drive them for ever away from her, the earth would have to impart a much larger velocity — an average of about twenty-six miles per second. The greater number of the expelled bodies would travel thenceforth on an orbit round the sun, crossing the earth's track at or near the place where they were first sent forth from their parent planet. One may almost say that this origin of many meteorites and meteor systems is forced upon us by the evidence. Still, it would be negatived if we found that volcanoes do not eject matter at all re- sembling meteorites in structure. The reverse, how- ever, is the case. Ranging the products of volcanic ejection in order according to the amount of iron they contain, and ranging meteorites in like manner, we find the two series coinciding over the greater portion of the longer — the volcanic series. We might L 2 148 OTHER SUNS THAN OURS. not, indeed, have known how closely the most fer- ruginous volcanic products resemble the iron meteor- ites* in structure but for the accident that Nor- denskjold discovered a mass which he mistook for an iron meteorite, but which is found now to be really a volcanic ejection, akin in structure to the field of basaltic lava (at Ovifak on the shores of Greenland), in the midst of which it had fallen while the lava was still plastic to retain this missile as it fell after its flight through many miles of air. We may, therefore, regard the terrestrial origin of many meteorites as highly probable, if not in effect demonstrated. Here Tschermak and Meunier pause, as also does Ball, who thus far had followed them. The last- named does not even ask, in that singularly interro- gative and unsatisfactory work, the " Story of the Heavens," whether we may not go further. For my own part, I find in this result the first step in a most interesting and suggestive path of inquiry. Regarding a large proportion of the material visit- ants of the earth as originally earthborn, we may conclude that in the remote time when our earth was a baby world, sunlike in condition, her path was traversed by hundreds of comets, her own progeny. These comets were followed severally by their trains of meteoric attendants. They were exposed to the action of those solar forces by which, within the last half century, a once promising member of another comet-family became dissipated until it finally lost altogether its cometic character. Millions of years ago, probably every one of them had been thus * It is worthy of notice that the Greek name for iron was derived from the name for a star, and iron meteorites are still called siderites. Before iron had been found in the earth, it was known only from the iron meteoritef v/hich had fallen at various times. Of course these were regarded as special pre- sents from the gods, and revered accordingly. WHENCE CAME THE COMETS ? I49 broken up until nothing remained but the streams of meteoric bodies, travelling round the orbit which had once been that of the earth-ejected comet. But this being the case with the earth, was the case also no doubt with every planet. Even our little moon,, whose scarred face still shows signs of the volcanic energies she once possessed, played her part in giving birth to such comets as she was equal to. If she possessed less volcanic power than the earth (at the same stage of the life of each), she required less power to eject matter for ever from her interior. On the other hand, the giant planets re- quired greater power ; but then they also possessed it. If Jupiter, for example, required power enough to eject bodies with a velocity of forty or fifty miles per second, yet it must be remembered that he is 310 times as massive, and therefore 310 times as strong as our earth. (For matter — inert matter as many choose to call it — measures in reality the strength of the orbs in space, and not only possesses power, but a power acting so swiftly across vast dis- tances that the velocity of light is rest by comparison. Moreover, this power possessed by " inert " matter is the source of every form of energy of which we know, even of life itself.) So with the other giant planets. Jupiter, then, and each one of his giant brethren, must during its sunlike stage have possessed the comet-ejecting power. Each giant planet must have had its comet-family at that remote time in the history of the solar system. And the comets thus formed by the giant planets, while no doubt very numerous, must, many of them, have been far more important than those to which our earth gave birth. Those comets would have lasted much longer, before dissipation due to solar disturbances set in. Then, also, the sunlike state of the giant planets must have lasted long after the earth and all the terrestrial ISO OTHER SUNS THAN OURS. ■ planets had passed that stage. For being so much' larger, the giant planets must have longer lives — the stages of planetary life being in effect stages of cooling. In fact, there are clear signs that neither Jupiter nor Saturn has cooled down to the earth's condition ; each is still too hot for the waters of its future seas to rest on its fiery surface. On this ac- count also, then, we might expect to find that some comets, sprung from giant planets and forming their families, might have remained even to the present time. Turning to the solar system, we find that this actually is the case. Nay, I myself, long before I had the least thought of attributing comets to plane- tary eruptive energies, had described the comets which hang about the orbits of the giant planets as " The comet-families of the giant planets." Some of the members of these families are among those from which the association between meteors and comets came first to be known. The meteor-stream from which the star-showers seen on August lo and 1 1 (the Perseides) proceed is not indeed associated with a comet depending on the orbit of any known planet ; but the meteors of November 13 and 14 (the Leonides) are associated with a comet depending on the orbit of Uranus ; and the meteors of November 27 and 28 are associated with a comet depending on the orbit of Jupiter — Biela's famous comet. Of course the members of these comet-families are exceedingly old. How old they are we cannot tell ; but that they are very old indeed is shown by the way in which, while they are unmistakably asso- ciated with the paths of the several giant planets, their orbits yet diverge far enough from those of their respective planet parents to indicate hundreds of thousands of years of perturbing action, unless indeed in some cases we may suppose that not the slow perturbing action of bodies at a distance, but WHENCE CAME THE COMETS? 151 the very active influence of some orb coming very close to a comet may have shifted tha comet's path. So many of their orbits pass through the widely-spread zone of asteroids, that we may very well imagine occasional very close approach to one or other of these bodies, and consequently a con- siderable change of orbit. It was thus that Sir John Herschel for a time tried to explain the disappear- ance of Biela's comet ; " May it not," he said, " have got entangled in the zone of asteroids, and have had its course altered by the influence of one of these bodies .' " Encouraged by the confirmation of the expulsion theory of comets, which we have found at this our first step, may we not boldly proceed yet one step farther — a long step I admit, but yet one suggested by the theory itself with which we are dealing .' The stars, like the giant planets, should have their part to play — a grander part of course — in the world of comet expulsion. They differ only from the giant planets, nay from the earth herself, in being in a different part of their orb-life. It is probable, in- deed, that among the stars there are orbs differing much less from Jupiter or Saturn than either of these still hot and fiery planets differs from the earth. Of course an orb like our sun, the one star we are able to examine, will require much greater energy to expel from his interior a flight of bodies, to become pre» sently a flight of meteors or a comet, than would a planet even of the giant type. Our sun, for example, would have to impart a velocity of 382 miles per second to a body ejected from his interior, that that body should pass away from his control for ever. But the sun possesses the required power. His mass (and therefore his might) exceeds that of the earth more than 320,000 times, that even of Jupiter 1,048 times. We have no means of recognizing by its orbital 152 OTHER SUNS THAN OURS. motion a star-expelled comet or meteor flight. But we need, not seek for bodies to tell us of expulsion, ages and ages ago. The stars are now in their sun- like state. They must therefore be doing such work now, if there is any truth in the theory to which we have been led. Now there is one of the stars which is near enough to be asked whether it really possesses and uses such expulsive power — our own sun. His answer is unmistakable. In 1872 and at sundry times since, he has been caught in the act of ejecting bodies, probably liquid or solid, through the hydrogen atmosphere around his globe, with velocities so great that the matter thus expelled from his interior can never return to him— the velocities ranging to45omiles per second at the least. What he is doing now he has doubtless done for millions, nay for tens of millions, of years in the past. What he has thus done, his fellow-suns the stars, thousands if not millions of mil- lions in number, have doubtless done also. Uncounted billions, then, of ejected meteor flights or comets must be travelling through interstellar spaces, visiting system after system, flitting from sun to sun, in periods to be measured by millions of years. The answer, then, to the question Whence came the comets ">. would appear to be : — (i) Comets which visit our system from without were expelled millions of years ago from the interior of suns. (2) Comets which belong to our system were mostly expelled from the interior of giant planets when in the sunlike state, but a small proportion may have been captured from without. (3) The comets of whose past existence meteor- streams tell us were for the most part expelled from our earth herself when she was in the sunlike state, but some of the more important were expelled from the giant planets, and a few may have been expelled from suns. CHAPTER XL A NEW THEORY OF SUN-SPOTS. Of all the phenomena presented to the contempla- tion of astronomers, sun-spots are at once the most impressive and the most mysterious. On the face of that resplendent disc they seem, at a first view, mere dark marks, of little import or interest. To the astronomers who first observed them, Fabricius, Scheiner, and Galileo, they were mere stains on the surface of an orb which earlier astronomers, confi- dent in half-knowledge, had regarded as absolutely without spot or blemish. But so soon as their real features are noted, and the real dimensions of the sun's orb considered, their amazing significance is revealed ; while, when their movements are ex- amined, and the strange laws noted according to which they wax and wane in frequency, they are found to present problems as mysterious as they are fascinating. I am about to advance a theory about sun-spots, or rather about their more salient features, which at least serves, whether right or wrong, to associate together some of the most remarkable facts which have been discovered respecting the sun and his surroundings. Let us first consider the nature of that surface in which sun-spots make their appearance, and the phenomena which they present. We are apt to regard the visible surface of the IS3 154 OTHER SUNS THAN OURS. sun as if it were either the actual surface of his globe, or, at least, very near to that surface. On a little consideration, however, of the facts known to us, it will appear that this view is not correct. Strangely enough, the earth under our feet tells us the nature of the interior constitution of the sun, while the face of the sun himself even veils from view what lies deep down below it. The crust of the earth, studied by geologists, has spoken in the clearest terms of many millions of years of sun work at the sun's present rate of emitting heat and light. We may shorten our estimate of the time by assigning to the sun a greater activity in past times than now, or lengthen it by assuming that of yore he worked less effectively ; but the result remains the same, so far as our present inquiry is concerned ; for it is the totality of sun work, not time, we have to consider. Dr. Croll, of Glasgow, has sho\(^n, if not conclusively, yet with such high degree of pro- bability that it would be far less safe to reject than to accept his conclusions, that the earth's crust tells of at least 100,000,000 years of sun's work. Sir Charles Lyell accepted the evidence as to all intents and purposes decisive. Yet if this is so, a great difficulty immediately presents itself. The sun's energy in emitting light and heat results, so far as can be seen, almost wholly from the action of gravity in drawing in towards the centre the matter which forms the great aggrega- tion we call the sun. . That mysterious power which resides in matter adds this other reason to the reasons, already strong, which make it the mystery of mysteries, that in it lies "the promise and potency" of light and heat throughout the universe itself We owe to Helmholtz the first suggestion and study of the theory which shows how the contraction of the sun's mass provides, so to speak, for the constant expenditure of energy. We can ascertain precisely A NEW THEORY OF SUN-SPOTS. 1 55 how much energy could have been derived from the contraction of the sun's globe to its present apparent size, supposing its mass strewn with tolerable uniformity through an orb of that size. Of course the larger the original volume of the sun, the greater the amount of energy which might thus have been produced. But let us assign to the original globe of the sun the greatest possible volume— infinity of space. Of course the idea is not admissible as a conception, but it can quite readily be dealt with mathematically, and will manifestly give us a superior limit to the length of time we wish to determine. We find, using this infinity of space, that the period deduced is but about 20,000,000 years. Taking, instead, an exten- sion all round over half the distance separating the sun from the nearest star, we get very nearly the same result. Here, then, there is manifestly something wrong. Our earth tells us one story, the sun seems to tell us another. I reject as absolutely inadmissible the suggestion for removing the difficulty by supposing that our sun's globe was formed by the collision of masses which had before been rushing with enor- mous velocities through space. All such ideas of collision appear simply preposterous to the astro- nomer who apprehends how enormously the dis- tances separating star from star exceed the dimen- sions of individual stars. There is only one way of removing the difficulty, viz., by recognizing the fact that the sun's apparent globe differs very much in size from his real globe. If the process of contrac- tion has gone on very much farther than it seems to have done, then we can readily explain the awful vistas of past time of which our earth's crust tells us. We may safely conclude from this one argument alone thaf the sun's real globe is verjr much smaller than the orb we see; IS© OTHER SUNS THAN OURS. But there is other evidence to the same effect. Professor G. H. Darwin has shown clearly that unless the central part of the sun were very much more compressed and dense than the parts near (say within fifty or a hundred thousand miles of) the apparent surface, there ought to be measurable flat- tening of the sun's polar regions. Now it is abso- lutely certain that there is no such flattening. All the observations made at Greenwich, Paris, Vienna, Washington, and other great observatories, agree in proving this. Therefore the central part of the sun is much denser than the outer parts, and doubtless the real globe of the sun is very much less than the globe we see. There is also another proof of the same important fact in the behaviour of the spots themselves. It will fall presently under our notice. What, then, is that visible surface which lies as a luminous veil far above the real surface of the solar globe ? The telescope shows the general surface of the sun as formed of multitudinous small round objects, intensely bright, on a background which, though really bright, appears by contrast dark. These objects are only small in the sense that they look small as seen even with the most powerful telescopes. In reality, they average two or three hundred miles in length and breadth. Regarding those of nearly circular form as in reality spherical, the surface of one of these clouds (if so we are to regard them), 200 miles in diameter, would be about 125,000 square miles ; so that in comparison with all such terrestrial objects as we can actually see and measure, they are of enormous size. Now we can readily form an opinion as to the nature of these cloud-like masses — the so-called solar rice-grains — by considering what the spectro- scope has told us about the vaporous atmosphere A NEW THEORY OF SUN-SPOTS. Ij/ in which they float. This complex atmosphere in- dicates its presence alike in telescopic survey of the sun and in photographs of his disc, by the well- marked darkening towards the sun's edge. Analyzed by the spectroscope, it is found to contain the vapours of iron, copper, zinc, aluminium, titanium, sodium, magnesium, and many other terrestrial elements, chiefly metallic. In other words, in the atmosphere of the sun the metals have the same posi- tion which the vapours of water have in our own air ; so intense is the heat of the sun that iron, copper, zinc, and so forth (doubtless, in reality, all the metals, though not all in sufficient quantity to indicate their presence), are turned to the form of vapour. The clouds, then, that float in the atmosphere of the sun are clouds in which drops of metal play the same part which drops of water play in our own clouds. We may describe the solar rice-grains, in fact, as mighty metallic clouds. But here I would call attention to a consideration which seems to me of great importance in all inquiries into the sun's condition. The laws of gaseous pressure and density, as determined by ex- periments on the earth, are either modified under the conditions which exist in the sun, or else we cannot possibly regard the region of absorptive vapours certainly existing around the visible surface of the sun as of the nature of an atmosphere. From spectroscopic analysis we know that the pressure at which hydrogen exists just outside the sun's surface is much below the pressure of our atmosphere at the sea-level, yet certainly not so low as the thousandth part of that pressure. And whatever opinion we may form as to the effect of the intense heat prevail- ing close by the sun, we cannot overlook the influence of the enormous force of gravity at his surface. Under this force, more than twenty-eight times the force of gravity at the earth's surface, an atmo- 158 OTHER SUNS THAN OURS. sphere constituted like our own would double in pressure for every one-eighth of a mile of descent. Suppose that at the sun's surface a vaporous atmo- sphere such as he seems to have, an atmosphere constituted as the vaporous matter around him un- doubtedly is constituted, doubled in pressure only once for every ten miles of descent. Then within the range of about 400 miles through which the sun's vaporous atmosphere has been observed (during total eclipse) to extend, there would be forty doublings, or the pressure, certainly not less than one-thousandth of our air's pressure, would be increased to more than one thousand million times the pressure of our air at the sea-level. Under such a pressure it would no longer be vaporous at all. Could it remain so, and obey the laws of gaseous matter, it would be many thousands of times denser than the densest metals known to us. Most assuredly no such pressure exists either at the sun's surface or thousands of miles below it. We can see to a depth of some 10,000 miles in the case of certain of the larger sun-spots. We seem forced to the conclusion that the real atmosphere of the sun does not come anywhere near the surface we see, which, according to this view, would be regarded as formed of clOud-like masses, each with its surrounding of vapour, kept around it by the attractive energy which must necessarily reside in enormous aggregations of metallic globules such as these clouds must be. I am aware that this view will seem so strange, so unlike any that has hereto- fore been held, as to appear very daring. Yet it is infinitely more daring to overlook the enormous physical difficulties involved in the assumption that a continuous atmosphere surrounds the sun to a height of many hundreds of miles, while at the highest part of that self-luminous atmosphere the pressure is comparable with that of our own atmo- sphere at the sea-level. A NEW THEORY OF SUN-SPOTS. 1 59 Be this as it may (for the question has no direct bearing on the theory I am about to present), it is certain that under the action of various forces the solar rice-grains arrange themselves into groupings of varied form, in such sort that the general surface of the sun, when studied with a telescope not suffi- ciently powerful to show the separate rice-grains, presents a mottled aspect Photography, which, as skilfully applied by Dr. Janssen, gives us the best views yet obtained of the details of the sun's surface, shows another reason for the mottled aspect, in the existence of a sort of network (varying even in form) of misty streaks where the rice-grains, though visible, are much less clearly defined than elsewhere. These blurred regions will doubtless find their explanation hereafter, as their changes of form come to be more closely studied. But, yet again, the surface of the sun is disturbed by forces producing more marked movements of the solar clouds. These get driven together into closely- packed streaks which, even in telescopes of very moderate power, are visible as exceedingly bright objects. They are the so-called /«i;«/^ (named thus by Hevelius), from the Latin word for little torches, because of their brilliant aspect. It is, however, when yet greater disturbances affect the cloud-laden region which forms the visible sur- face of the sun, that solar spots make their appear- ance. A region of disturbance, where many faculee are seen making the sun's surface look like a froth- streaked sea, shows suddenly in the middle of a dark region, round which the faculse appear at first as parts of nearly circular arcs. But they pass farther and farther away from the region of disturbance, the dark centre of which becomes better defined, and is presently seen to be bordered by a well-defined fringe of less darkness. Under close telescopic scrutiny this fringe (called the penumbra), which, though less l6o OTHER SUNS THAN OTIRS. dark than the central part (called the umbra), ts darker than the general surface of the sun, is seen to be marked by streaks extending radially from the centre of the nearly circular spot. Larger and larger the spot grows, gradually losing its circular form, but still well rounded on all sides. The centre is found to be darker than the rest of the umbra, ap- pearing, indeed, absolutely black, but not necessarily so, since the glowing lime-light appears absolutely black when on the sun's disc as on a background. This central darkest region is called the nucleus. After remaining, sometimes for several days, some- times for weeks or even months, a spot begins to show signs of breaking up, if one can speak of the breaking up of what really indicates the absence, not the presence, of matter. It loses its rounded form, becoming perceptibly pear-shaped. Large portions of the facular regions around break their way in upon the sun, chiefly on the edge, which re- mains more rounded, forming often bright bridges — usually curved — from side to side of the spot. On either side of the smaller part (the stalk end of the pear) larger but less brilliant masses seem to move in upon the spot, as though to cover it over with portions of the cloud-laden surface which had before been outside. These masses, as they move on, usually show widening dark streaks between them ; and it is very noteworthy that on either side of these dark streaks there can be seen bright threadlike ob- jects akin to the radial streaks around the umbra. But in the meantime these streaks, which had been originally radial and tolerably regular, have been tossed hither and thither as if irregular currents swept them in different directions. From the great masses thrown in on the dark background of the spot multitudinous filaments seem to stream in all direc- tions, like fringe upon a storm-tossed banner. More and more violently — "pell-mell," as Secchi A NEW THEORY OF SUN-SPOTS. l6l used to say — the luminous masses rush in upon the spot region. At last it is completely covered over, though bright facular streaks show where the great opening had been, and where intense disturbance is still going on. Sometimes these streaks break apart and a fresh spot is formed ; and it has happened that twice or thrice a spot has been, as it were, renewed in this way. But usually the facular streaks become less and less marked, until at length the region where the spot has been can be in no way distinguished from the surrounding parts of the sun's surface. Such is the history of a spot of the larger sort. Occasionally there are peculiarities affecting the pro- gress of some particular spot. For instance, there was the wonderful Cyclone Spot, seen by Secchi in 1857, the whole area of which was swept round as if by some mighty tornado. Again, there have been spots where a double tornado seems to have been in progress, the two whorls moving in opposite direc- tions. In yet other cases there has been a whirling motion affecting the central part of the spot region in one direction, at one part of the spot's career, and in the contrary direction later. Other evidences also of exceedingly violent motion have from time to time been observed. In smaller spots less marked signs of varying dis- turbance are noticed. The history of a small spot is comparatively uneventful. The chief interest in these lesser markings resides perhaps in the circum- stance that to the unpractised observer they look very much like small planets in transit. For my own part, I may express my conviction that every recorded case of intra-mercurial planets seen in transit is to be thus explained, from the case of Lescarbault's Vulcan down to the case of Vulcan's supposed return as seen in China ; though the last-named is the only case in which a photograph of the sun chanced fortunately to have been taken at the right time, proving unmis- takably that what had been described as unquestion- M r62 OTHER SUNS THAN OURS. ably a planet, moving like a planet and unlike a sun-spot, was nevertheless a small sun-spot after all. But there are yet some other circumstances which must be noted before we proceed to consider a theory of sun-spots. The spots are limited to two zones on the sun's surface, corresponding to the sub-tropical and tem- perate zones on the surface of the earth. The existence of such zones implies necessarily the occurrence of rotational motion, whereby the position of the sun's poles and equator has been determined. It has been, in fact, by observing the spots, that the axial position of the sun and his rate of rotation have been ascertained. But the movement of rota- tion, which seemed a comparatively simple matter when the first rough observations of Galileo and his contemporaries were in question, presents itself now as a complex phenomenon ; for spots' in high solar latitudes are found to indicate a rotation rate different from that determined by the observation of spots near the equator. The difference is so great as to become most perplexing when its real significance is considered. Judged by spots in the highest latitudes where spots have been seen on his face, the sun seems to rotate in about twenty- eight days. Judged by spots as near the equator as any have been seen, he seems to rotate in about twenty-four days. His real globe cannot well rotate save as a whole and in a single period ; yet, judged by what looks like his surface, his equatorial regions seem to rotate seven times, while the mid-zones of his northern and southern hemispheres rotate only six times. Re- garding the slower rate for a moment as the true rate of the sun's rotation, it would appear as though the visible equatorial regions gained one entire rotation on the surface beneath them in i68 days. Now, the sun's circumference is in round numbers about 2,660,000 miles, so that the gain of the whole A NEW THEORY OF SUN-SPOTS. 163 equatorial zone takes place at the rate of nearly" i6,cx)0 miles per day, or about 650 miles per hour. Thus, viewing the varying rotation rate at the sur- face, we should have to recognize the existence of the most stupendous and far-ranging hurricane the mind can conceive. We may fairly find in this amazing mobility an- other and simpler proof of what we have already seen to be demonstrated by subtler evidence, the vastness of the distance which separates the real surface of the sun from that visible surface which we call the photosphere. One other point remains to be mentioned. The spots, besides being limited in space, are limited also in time. They cannot always be looked for with any probability that they will be seen. At this present time there are many spots on the sun's face. But if he is watched week after week during several coming years, it will be found that the spots grow fewer and fewer till none are seen. Then several weeks, or mayhap months, will pass during which no spots and few faculsE will be seen, when the mottling will be scarce discernible, and the darkness near the edge will be much less marked than usual. Then the spots will begin to return, will become more and more numerous till they attain their maximum fre- quency. Then they will diminish till they disappear, then return, then pass away again, and so on con- tinually, waxing and waning with a sort of rhythmic flow. But the oscillation is not uniform. The average interval between two successive epochs of greatest ' spot frequency is a little greater than eleven years, but the interval has been as short as eight years, and it has been as long as sixteen years. Such being the most striking peculiarities of the sun-spots, let us see whether they can be associated together, some or all of them, by any theory as to M ^ 164 OTHER SUNS THAN OURS. the way in which these great openings in the lumin- ous cloud region are formed. In the first place, it may be fairly assumed that the real seat of the disturbance seen when a spot appears, lies below the visible suiface of the sun. There are, indeed, several circumstances which seem at a first view to suggest that the disturbance has its origin from outside. If the spot period were of con- stant length, one might be led to suppose that some as yet undiscovered comet, having a period of about eleven years, and followed by a train of meteoric attendants, travels in an oval orbit intersecting the outlying cloud envelopes of the sun, and periodically with its flight of meteoric followers breaks through the region of luminous clouds. There are also cer- tain peculiarities of sun-spots, noted by the late Mr. Richard Carrington, which have been held to indicate an external origin. But as none doubt that the real energies of the sun reside in that concealed mass which lies within the photosphere, hidden by a veil through which man can never hope to penetrate, and as the spots by their size and movements tell of most energetic disturbing forces, we must, it would seem, look for their origin where alone such forces are at work. Again, if the origin of the spots is below the photosphere, and at the real surface of the sun, since the distance between this surface and the photosphere is enormous, we can hardly imagine any way in which forces exerted at the surface can affect the photospheric cloud region, unless they are directed with great energy radially from the sun's surface. In other words, it would seem that the forces at work in producing sun-spots are eruptional. Now, if we conceive the outburst of masses of strongly compressed and intensely heated gases from below the sun's real surface, and trace the result of their uprush, we are led to recognize certain pheno- A NEW THEORY OF SUN-SPOTS. 1 6? mena, which certainly correspond well (be this ex- planation true or not) with what is seen on the sun. Even if the theory is incorrect, it has its value in thus associating together, as will be found, the various facts known about sun-spots, the coloured flames, and the solar corona. Let us suppose that a great eruption begins deep down below the visible surface of the sun, imprisoned gases bursting their way forth, and in their outburst driving masses of solid or liquid matter like missiles through the distant photosphere. As the compressed vapours travel onwards to regions of diminishing pressure, they would expand, cooling in the process, and drive away from all round the region where they reached the visible surface the clouds which had covered that region. At the beginning there would be a central space, from around which the clouds were thus cleared over a continually widening area. Moreover, regarding the visible surface as part of a cloud stratum of great thickness (certainly not less than 10,000 miles in depth), it is clear that the con- stantly expanding masses of vapour, in their upward rush, would drive the higher parts of the cloud region farther apart than the lower portions. Thus looking squarely into the opening, from outside, as when we look at a spot near the centre of the sun's face from our terrestrial standpoint, we should obtain slant views of the cloud stratum. Now, the clouds which had before been spread uniformly over the scene of disturbance, being driven away from it upon the surrounding region, would necessarily be packed closely together, and so would form luminous streaks all around the spot — the faculae, which, as we have seen, surround the dis- turbed region. The penumbra would show what lies underneath the photosphere, but not in its normal condition ; for the mighty uprushing and side-thrust- ing masses of vapour would displace all parts of the 1 66 OTHER SUNS THAN OURS. cloud stratum, even as the outer parts are displaced and made to form facular streaks. Still we can form an idea, from the aspect of the penumbral fringe, re- specting the normal condition of the inner parts of the solar cloud region. The radiating streaks, which are manifestly slant streaks of luminous matter below the clouds, seem to tell us clearly of streaks which had been vertical before the disturbance. We may compare what we see round the spot to what one would see in looking down upon a field of wheat (from a balloon, suppose) over a part of which a small but violent whirlwind was passing. All round the centre of disturbance the stalks of wheat would be driven aslant, and we should see them sloping radially around that centre. The ears of wheat be- longing to the storm-bent stalks would be driven closer together than the ears elsewhere over the field, and so would form circular streaks around the region of disturbance, and outside the slant radial streaks. These circular streaks of compressed wheat-ears would look brighter than the rest of the field if the ears were in their golden prime. So the glowing solar clouds, urged together by the expansive action of the vapours poured into the spot region, form streaks looking brighter than the surrounding sur- face ; while extending from them inwards, towards the spot's centre, are seen the streaks of luminous matter which before had been vertical. What these vertical streaks may be is not very easily determined. They may be down-rushing streams of molten metal from the sun's metallic clouds, or they may be up- rushing columns of glowing metallic vapours, capped by the clouds (as in our own air, uprising streams of aqueous vapour are capped by cumulus clouds), or they may include both forms ; however they are to be interpreted, it is certain they exist. After a while the eruptive forces cease ; the ejected vapours for a while continue to extend themselves A NEW THEORY OF SUN-SPOTS. 167 around the region of disturbance, but not long. All the forces now called into action are such as tend to fill in again, and cover over, the region which had been disturbed. As the surrounding cloud-covered regions strive to rush in, contests arise between the in-rushing masses and the vapours within the spot region. In these conflicts cyclonic action may arise, and usually does. Sometimes a single cyclonic whirl is generated ; at other times two or more, which may be in the same or different directions ; while at yet other times, changes in the conditions under which the conflicts take place, may cause a cyclone in one direction to be replaced by another in the contrary direction. Again, the enclosed vapours would maintain a better resistance and preserve the rounded form of the spot on that side towards which their motion urged them. On the other side, where the resistance would be less effective, cloud-laden masses from the solar photosphere would break in, or rather would be drawn in ; and around this part of the disturbed region the photosphere would be more disturbed than elsewhere, and in many parts would be broken up. The masses thus flung over or projected towards the region of the spot would be agglomerations of the luminous clouds with their vaporous surroundings and their filamentous appendages, which, in the more quiescent parts of the sun's surface are usually (it may be presumed) nearly vertical. A mass of clouds driven onwards, as by a mighty but irregular hurri- cane, would show its filaments as streamers from a wind-tossed pennon, as these luminous thread-like forms actually appear. Not parallel here, as around the edges of a yet youthful spot, the filaments would present an appearance more nearly resembling that of our cirrus clouds, with their wild mare's-tail streaks tossed seemingly hither and thither by the varying currents in our upper air. Indeed, Professor l68 OTHER SUNS THAN OURS. Langley, to whom we owe decidedly the best views of the various features of the sun's surface yet drawn, finds every form of solar cloud illustrated in the clouds of our own air. But though we may thus find illustrations of solar features, we must not imagine that therefore we have necessarily their true analogues. The vast difference of scale must be carefully kept in recollection. The solar clouds, which seem simple rounded masses of lumi- nous matter, are in reality vast cloud balls, two or three hundred miles in diameter ; and doubtless, could we see them more clearly, would show amazing irregularities of structure where our present tele- scopes show uniformity. The filaments merely look like the thread-like forms which we see in our cirrus clouds ; in reality they are forty or fifty miles in breadth, and some of them are fully 10,000 miles in length. Nothing that we know about our clouds enables us to form the merest guess as to the con- dition of such vast masses, such long streamers as these, or even to say that they are single masses or continuous streamers at all. And apart from all this, the intense heat which pervades the whole material of these seeming clouds and seeming streamers assures us that they are as unlike our clouds and cloud streamers in condition as they are in volume. All that we can here say is that the sun-spots behave as though they were produced by the uprush of masses of vapour, caused by eruptive action far below the visible surface ; for all the phenomena presented by a spot, from its first formation to its final disappearance, correspond with what might fairly be expected to result from such a process of forma- tion. In passing, however, it may be noted as strong evidence in favour of the theory that sun-spots are due to the action of forces working below the visible surface, that they are regions of darkness and not of A NEW THEORY OF SUN-SPOTS. 1 69 increased brightness. If sun-spots are produced in the way I have suggested, there would result great cooling from the expansive action of vapours which had been enormously compressed. On the other hand, if sun-spots had their origin from without, the bringing to rest of matter, meteoric or cometic, which had before been travelling with enormous velocity, would necessarily be accompanied by the generation of heat. Since the spots by their darkness and by the spectroscopic evidence of powerful absorptive action tell us that they are regions of cooling and not of greater heat, we may reasonably and safely infer that they are due to the action of forces working from within expansively, and not from outside with effects of compression. But now let us see whether we may not find other evidence bearing on this theory of sun-spots, by looking outside the sun's surface for the effects, even as we have looked below for the cause of the dis- turbance to which they are due. So soon as the coloured prominences had been shown by Lieutenant (now Colonel) Herschel,Janssen, Rayet, and others, to be great masses of glowing gas, it became possible to observe them without waiting for total solar eclipses. Shining with special tints only, their light could, by spectroscopic dispersion, be brought into rivalry with only such light from the surrounding sky, or even from the sun himself, as is of one of those tints. The totality of sunlight over- whelmingly surpasses the totality of prominence light ; but red light from a prominence is not over- whelmingly surpassed by the red light of the same or very nearly the same tint, either from the sun or from the sunlit sky close by him. Thus, by keeping out all light save that of this special red tint of hydrogen, or if preferred the orange-yellow tint of helium, or either the indigo or the greenish-blue tint of hydrogen, the shapes and movements of lyo OTHER SUNS THAN OURS. the great coloured flames can be discerned and watched. Now the most interesting of all the results which have followed from the application of this fertile method of observation has been the division of the coloured prominences into two definite classes. First there are the cloud-like prominences, which in form and movement closely resemble the clouds of a wind-swept sky, or sometimes of a sky comparatively calm. Secondly, there are the jet-like prominences, which by their form (their initial form at any rate) and by all their movements show that they are due to eruptive action. The cloud-like prominences appear around all parts of the sun's edge, which is equivalent to say- ing that they occur at all parts of the sun's surface. In this respect they are like the solar clouds and the faculae. They are apt to be somewhat larger and more numerous opposite the spot-zones, which amounts to saying that they occur in greater relative frequency, and attain a greater average size, over the spot-zones. In this respect they resemble the faculae. It seems likely therefore that if (as is most probable) there is some connection between the coloured prominences and the phenomena of the sun's surface, the faculae are the features to be specially associated with the coloured prominences of cloud-like form. These cloud flames often attain an enormous size and height, reaching sometimes eighty or even a hundred thousand miles above the sun's surface. They are less brilliant than the eruptive prominences, and though their movements (or rather their apparent changes of form) are some- times amazingly rapid when compared with the movements of terrestrial clouds, yet they show nothing like the rapidity of motion observed in the prominences of jet-like form. The cloud flames may be looked for at all times, whether the sun shows A NEW THEORY OF SUN-SPOTS. I/I many spots or few, or none ; but they are apt to be rather more numerous when there are many spots. The eruption prominences, on the other hand, are never seen except opposite the spot-zones, or, in other words, they never exist except over these zones of the sun's surface. Moreover, the jet prominences are only seen when there are spots on these zones ; and though this has not yet been actually established by observation, there are strong reasons for believing that an eruption prominence is never to be seen ex- cept above a solar spot. Their occurrence only over the spot-zone, and at a time when there are visible spots, suffices of itself, however, to prove that they are intimately connected with the occurrence of that particular kind of disturbance which results in the breaking up of the photosphere and the formation of sun-spots. This being so, it becomes probable, on d priori grounds, that by studying the jet-like prominences we may obtain information about sun-spots ; and vice versd, that any true theory we may be able to form respecting sun-spots will throw some light on the nature of the eruption prominences. These jet-like protuberances are generally smaller, brighter, and better defined than their cloud-like brethren. They have usually been regarded as actual eruptions of glowing hydrogen ; but this view seems as incorrect as would be the idea that the smoke and products of chemical action flung from the mouth of a cannon are the real missiles ejected. We may, indeed, by noting the behaviour of the glowing hydro- gen in the eruption prominences, obtain clear and decisive evidence that it is to the smoke from a cannon they are to be compared rather than to the ejected missiles. We see lofty columns of the glow- ing hydrogen at first as though they had themselves been flung forth as mighty streams of gas from the sun's interior ; but a few minutes later the upper t72 OTHER SUNS THAN OURS. parts of these columnar streams spread themselves out into cloud-like forms, much as the smoke which at first rushes straight enough from the mouth of a cannon begins presently to expand into cloud-formed masses. Such, for instance, was the behaviour of a mighty spiral column of glowing hydrogen seen by Zollner as far back as 1870, and pictured in my treatise on the sun. Here was a column 32,000 miles in height, so that four globes like our earth, placed one upon the top of another, would not have reached to the summit of that long column. How unlikely, on the face of things, that a rare gas such as the hydrogen then seen (for, by the spectroscopic method of observation, its density could be deter- mined and was found to be small) could be ejected through resisting vaporous matter to so enormous a height. But even could this have happened, it is certain that after rushing thus far the hydrogen would continue to ascend in jet-like form, not begin to spread into cloud form just where the jet-like motion would have become possible in consequence of the greatly diminished resistance. If any doubt could remain after the consideration of such cases, it would be removed by the phenomena presented during the remarkable eruption witnessed by Professor Young in 1871.* On that occasion a long low-lying cloud of glowing hydrogen was torn into shreds by a tremendous outburst which occurred below. Long filaments of hydrogen were seen travelling upwards so swiftly that their motion was actually discernible, a circumstance very unusual, and meaning a great deal at the sun's distance. Higher and higher these filaments of hydrogen seemed to rush, until at. last they had attained the enormous height of 210,000 miles (at least)f from * Eruptions of a similar character have been witnessed since, but that was the first that had ever been seen, t They may iiave passed much farther away than this, for A NEW THEORY OF SUN-SPOTS. 1 73 the sun's visible surface. Even at that enormous height they did not cease to ascend ; they simply lost their lustre and became no longer discernible. From a calculation based on the observed time in which this enormous distance seemed to be traversed, I determined the velocity with which the matter ejected on that occasion crossed the visible surface of the sun at certainly not less than 300, and pro- bably not less than 500 miles per second. Now the filaments of glowing hydrogen by no means pre- sented the appearance of bodies rushing with enor- mous velocity through a resisting atmosphere. On the contrary, they were long irregular streaks of luminous gas, pointed in front (with reference to the direction of their motion) as well as in the rear. I do not think they can possibly be regarded as the missiles then ejected. Their motion was probably apparent only, not real. I take it that when one of these filaments was seen apparently advancing with enormous velocity upwards, what was really happen- ing was this : A solid or liquid mass was rushing upwards, tearing its way through whatever hydrogen lay along its track, and thus leaving behind it a trail of glowing hydrogen, growing at the upper end as the missile advanced, and losing length at the rear end as the imparted heat passed away, and so appearing to advance — even as the trail of a meteor seems to advance ; though in reality the luminous matter forming that trail has not passed onwards, but the meteor passing onwards has caused atmo- spheric regions continually farther and farther for- wards to become luminous. It is tolerably obvious that on this occasion there was an ejection of matter solid or liquid (or if the distance measured was the apparent distance ; and if their course was aslant to the direction of the line of sight, the real distance was certainly greater, and may have been much greater. J 74 OTHER SUNS THAN OURS. vaporous, then of great density) at velocities so great that the ejected matter could never return to the sun. A velocity of about 380 miles per second is the greatest the sun can control in matter at his surface. In this case the ejected matter probably crossed the sun's surface with a velocity far exceeding this, and is now travelling, with velocity constantly diminishing but never to be entirely lost, into the remote depths of interstellar space. It is difficult to see how so enormous a velocity as this could have been acquired or imparted below that mobile surface which we call the photosphere. Professor Young has suggested that the sun is a gigantic bubble, and that beneath the skin (really the enclosing strata) of this bubble the forces of outburst may be restrained until they acquire the energy necessary to expel matter at the observed rate of ejection. But every- thing in the behaviour of the great eruption promi- nences speaks of an origin much more deep-seated than the inner layers of the photospheric cloud regions. Doubtless it is at and below the real surface of the sun that the eruptions occur by which missiles are ejected through the solar cloud envelopes, to pass in some cases but a few thousand miles higher, in others hundreds of thousands of miles away through the heart of the corona, and in yet others beyond the very limits of the solar system itself* Lastly, in the corona itself we find evidence of the action of eruptive or repulsive forces in the solar spot region, though indirectly rather than directly. There is, indeed, direct evidence of some such action * It is noteworthy that in 1864 Mr. Sorby, of Sheffield, was led by the microscopic study of meteors to the belief, or rather to the conviction, that they had once been either in the interior of our sun, or of a body in the sunlike state ; while the late Professor Graham, of London, was led to a precisely similar conclusion respecting the Lenarto iron meteor, by the quantity of hydrogen which he found occluded within its mass. A NEW THEORY OF SUN-SPOTS. 1/5 in the greater extension of the corona opposite the spot-zones. But the indirect evidence is stronger. The light of the corona, under spectroscopic analysis, is found to be partly reflected sunlight, partly in- herent light due apparently to two sources ; — first, incandescent solid and liquid matter in the neigh- bourhood of the sun, and secondly glowing gas. The lines of glowing hydrogen show that this gas is present in the corona at times, if not always, though assuredly not as the component of a gaseous atmo- sphere extending from the sun to the distance of even the inner bright corona. But it is noteworthy that the lines of hydrogen have only been seen or have only been bright at a time when there have been many spots on the sun's face, and therefore at the season when eruption prominences appear. It seems reasonable to infer that at such times the eruptive or repulsive action of which the jet pro- minences give evidence, leads to the ejection or repulsion of meteoric and cometic matter through the hydrogen present in the corona, and conse- quently to the heating of the hydrogen in such degree that its bright lines show under spectroscopic scrutiny. It seems certainly noteworthy that so many phenomena presented by the sun-spots themselves, the coloured flames, and the corona, accord so well with a theory originally advanced only as a suggested way of interpreting certain features of the solar spots. Whether the theory is sound or not, it serves con- veniently to associate a number of highly interesting facts respecting these phenomena of the sun and of sun-surrounding space. CHAPTER XII. TWO SUNLIKE PLANETS. In my last essay, in presenting "A New Theory of Sun-spots," I point out certain considerations which involve in reality changed ideas as to the sun's actual condition — but ideas not altogether new, since they were propounded by me several years since. According to these views, the surface of the sun, as we see it, is simply the outside of a region of clouds, having a visible depth of some ten thousand miles, and separated from the sun's actual surface by a region of vaporous matter probably many tens of thousands of miles in depth. I point to three kinds of evidence in favour of this view, in fact absolutely demonstrating its justice : first, the varying rate at which the sun-spots are carried round, those near the equator completing seven circuits while those farthest from the equator com- plete but six ; secondly, the evidence of our earth's crust, which tells us that the sun has been at work far longer than we could infer if his real globe were as large as the globe we see ; and thirdly, certain mathematical calculations by Professor G. H. Dar- win, which show that unless the sun's central parts were much more compressed than the rest there would be measurable flattening at his poles. The consideration of the vast distance thus shown to separate the real surface of the sun from the surface 176 TWO SUNLIKE PLANETS. 1/7 we see, led to the inference that the cloud region which constitutes his apparent surface can hardly exist within an atmospheric envelope properly so called, simply because within such an envelope the pressures would increase so greatly with approach towards the sun's centre that within much less than the hundredth part of the distance separating that surface from the cloud region, pressures changing the gaseous into liquid matter must inevitably arise. Then we were led to consider the evidence given by the sun-spots, prominences, and corona, finding reason to believe that the sun-spots are phenomena of eruption from beneath the real surface of the sun, that in the mighty eruptions producing these phenomena matter is driven out through the region of the coloured flames, outwards even through the whole coronal region, farther yet to the very out- skirts of the solar system ; nay, even in some cases beyond the limits of the solar system into the inter- stellar regions. Now it is becoming generally recognized that in suns and planets, in all the orbs in fact which people space, there are stages of existence akin to the stages of life. There is a period of preparation, a period of youth, a period of mid-life, a period of decay, and finally there comes the end of life. The stages of an orb's life may be described as, in the main, stages of cooling. In suns we find evidence of such changes in the different condition (as shown by the spectroscope) of such orbs as Sirius, Vega, and Altair on the one hand, and orbs like our own sun, much smaller and therefore much more advanced in orb life, on the other. But also we find much older suns than ours in the orange-red and yellow stars, and older orbs still in the blood-red and garnet- tinted suns shown by the telescope in various parts of the star-depths. Nor do astronomers doubt that there are older suns yet, suns which have passed N 178 OTHER SUNS THAN OURS. on to the last stage of all (mayhap), the stage of darkness. Moreover, as it is a known law of cool- ing bodies that the larger a cooling mass of given temperature is, the longer will be the stages of its cooling,* we may safely assume that, apart from differences (which may nevertheless be enormous) in the time of beginning orb life, the larger suns, hav- ing much longer periods of life, will have passed through less of their longer lives than the smaller suns. If we extend such considerations to our own solar system, as indeed we may do with much more confidence, since, forming a single system, it has doubtless had a simple history, we find certain very interesting ideas suggesting themselves in relation to the various orders of bodies forming that system. We see that the giant planets, for example, being very much smaller than the sun, must have much shorter lives. The sun exceeds Jupiter 1,047 times in mass, and Saturn is less than a third even of Jupiter. It is clear that, even granting the sun a start of millions of years of orb life as compared with these giant members of his family, he would still be very much younger than they are in develop- ment. We understand, in fact, how ' it is that, whereas he is in the sunlike or glowing vaporous stage, they no longer have the sunlike aspect. But, on the other hand, they exceed our earth so enor- * The law is experimentally verified in a great number oi well-known cases, but the reason is not far to seek. The quantity of heat in a mass of matter, at a given temperature, is of course proportional to the mass (comparing bodies of the same constitution), which varies as the cube or third power of the linear dimensions, whereas the rate of parting with the heat is necessarily proportional to the surface whence alone the heat can pass, that is, to the square of the linear dimen- sions. Hence, in bodies of the same substance and similar in shape, the duration of a process of cooling is proportional to the linear dimensions. TWO SUNLIKE PLANETS. I/g mously both in size and in mass, that by parity of reasoning she ought to be very much older than they are (in development always, I mean, not in years). Jupiter exceeds her 310 times, Saturn exceeds her ninety-seven times in mass, and such differences as these imply not only difference in degree but kind. Jupiter and Saturn must not only be more youth- ful than the earth, but in a different stage of orb life altogether. We may say that they must be in a stage intermediate between the sunlike and the earthlike — they may be expected to show under careful study evidence of conditions alien to those found in the sun, as probably at least as of con- ditions resembling those existing in the case of our earth. It seems to me that, this being so, we may reasonably look to the giant planets to give evi- dence respecting the sunlike stage of orb life, to show features such as we have recognized in the sun — however readily we admit, as of course we must, that they are not sunlike now. At the very outset of the inquiry we find a re- semblance which is at least striking, even if it be merely one of those which are rather accidental than actually significant. Jupiter's system is an almost perfect miniature of the central part of the solar system. If for a moment we regard Jupiter and the other outside giants of the system as not really forming part of the sun's family, then that family would consist of four worlds (one of them double) : Mercury, Venus, the Earth-and-Moon, and Mars. Jupiter similarly has a family of four worlds, his four moons, and the paths of these lie at dis- tances closely corresponding, but on a smaller scale to the distances separating the paths of the terres- trial planets. Moreover the moons of Jupiter are by no means such insignificant bodies as their tele- scopic aspect might seem to suggest. The least of N 2 l8o OTHER SUNS THAN OURS. them has a surface as large as that of North and South America together, — the largest is not much smaller than the planet Mercury. In fact the one feature which spoils the perfection of the miniature picture formed by the Jovian system is that his moons are relatively very much larger than the planets. Again, in the case of Saturn we have a system of eight worlds, the largest of which is nearly as large as Mars, the second in size nearly as large as Mercury, while the least has a surface large enough to be the abode of many millions of living creatures. We thus see in Jupiter and Saturn, as in the sun, orbs ruling in one sense, orbs serving (but generously) in another sense. Each bears potent sway over a family of worlds, but each pours forth in plenty rays of light and heat by which life on those worlds may be nourished. This is true even if we have to regard Jupiter and Saturn as only reflecting rays of solar light and heat ; simply because their power to so de- flect towards their subordinate worlds the rays sent forth by the supreme centre is so enormous. I have calculated that in the skies of his nearest moon Jupiter must show a disc 1,300 times as large as that of the full. moon. In the heavens as seen from the innermost Saturnian satellite, the ringed planet must appear so enormous that when the outer edge of his ring system touches the horizon the opposite part of that outer rim must reach to the point overhead.* * It may be remarked in passing how carelessly writers on matters astronomical have accepted and repeated the sugges- tion of Brewster, Whewell, Chalmers and others, that the four moons of Jupiter make up to that planet for the small amount of sunlight he receives from the sun, while the eight moons and the ring system of Saturn play a like part, with even greater effect, for that still remoter planet. As a matter of fact, the total amount of sunlight reflected by all the moons of Jupiter is barely the sixteenth part of that reflected by our full moon ; and the eight moons of Saturn reflect rather less. As for the rings of Saturn, I have shown in my treatise on Saturn TWO SUNLIKE PLANETS. l8l Here, then, at once is a solar attribute possessed by the giant planets — unlike the earth, they are like the sun in being the centres of systems of circling worlds. So far as analogy can be our guide at all in such matters, and I must admit it is not an altoge- ther trustworthy guide, it would seem that we should rather regard Jupiter and Saturn as orbs nourishing life in a system of circling worlds, than as them- selves ^yorlds fit to be the abode of multitudinous forms of life. But now let us turn to the more trustworthy evi- dence afforded by physical features — let us see what the telescopic aspect of the giant planets, and what the spectroscopic analysis of their light, may suggest as to their actual condition. Here again, however, we are immediately struck by solar rather than by terrestrial features. We see a surface of cloud, not a surface of land and water. We see evidence of enormously rapid rotation. We recognize the existence of parallel banks or belts of clouds, akin in some degree to the parallel zones in which the sun-spots travel. When we watch the movements of the markings on the belts, we find that they are not only carried round by the rota- tional motion, but have different rates of motion, indicating (much as we found in the sun's case) widely different rates of rotation in the various zones of the planet. Now the interpretation we were forced in the sun's case to place upon the existence of different rates of rotation in the visible surface, was that the real sur- face of the sun lies very far below the surface of and its System, by mathematical demonstration which there is no disputing, that they serve, not to add to the planet's supply of sunlight, but enormously to reduce it, actually casting large parts of the planet into total eclipse, lasting for five or six of our years at a stretch, and causing scarcely less disastrous eclipses to every part of the planet from which the rings can be seen at all. 1 82 OTHER SUNS THAN OURS. clouds which we actually see. We found other evi- dence which not merely supported but actually demonstrated this view. Let us see how the case stands with Jupiter. Is there any other evidence to confirm the belief that the visible cloud-surface lies at an enormous distance above the surface we see .' It appears to me that, although in this case we want the evidence which in the sun's case we derived from the crust of the earth, we have very strong evidence to show that the real globe of Jupiter is very much smaller than the globe we see and measure. Compare first the quantity of matter contained in Jupiter with what we should infer from his apparent size. He is 1,250 times as large as the earth, but only 310 times as massive. Yet every part of that great mass of his possesses the power of attraction to compress the planet's substance towards the centre. Made, as in all probability* Jupiter is, of the same materials as the earth, we might fairly ex- pect him to be a much denser rather than a much rarer planet. Even if his whole mass were molten through intensity of heat, still we might expect the slight expansion so arising to do little more than counterbalance the effect of the enormous self-con- tracting power residing in his mass or weight, even if it did as much. We are justified, then, seeing his mean density is but one-fourth of the earth's — or, as it chances, almost exactly the same as the sun's — in inferring that he is not so large as he looks. Doubt- less his real globe is at the very least as dense as the earth's. In this case the volume of the true Jupiter is but one-fourth the volume of that globular * Since the central sun is of the same material as the earth — one of his family^we may infer that he is of the same material as the other members of his family (for why should the earth differ from the rest ?). If so, it follows that all the members of the solar system are formed of the same materials. TWO SUNLIKE PLANETS. 183 space, enclosed within vast cloud-layers, which we measure and regard as the real globe of the planet. This would assign to the true Jupiter a diameter of less than two-thirds his measured diameter, or, making his radius about 26,000 miles instead of about 40,000 miles, would leave a distance of at least 14,000 miles intervening between the surface of the real globe and that outside surface which we see and measure. Next take the case of Jupiter's brother giant, Saturn. Here we have apparently an even younger orb than Jupiter. Saturn's ring system is in reality a part as yet unfinished of his system of dependent bodies. It consists of multitudes of tiny bodies travelling in the same general plane, and like sands on the seashore for number. Hereafter, under the mighty forces of the planet's energy of attraction, this system of rings will be broken up to form two or three other worlds akin to the eight satellites which already travel round the planet. While we thus find evidence of extreme youth in the ring system, we find confirmation in the singularly small density of Saturn. With a volume exceeding that of the earth seven hundredfold, he has less than 100 times her mass. We must explain this in the same way as in Jupiter's case. We must suppose Saturn's real orb to be more than a hundred times the earth's globe in volume, that is, one-seventh part of the volume of the cloud-enwrapped space we measure as if it were Saturn's veritable globe. This would make the diameter of Saturn fully 16,000 miles below the surface we see and measure ; or, taking the mean radius of his cloud-surface at 36,000 miles, his actual radius would be about 20,000 miles. Observe, now, the evidence of those parallel belts into which the cloud-surface both of Jupiter and of Saturn is nearly always arranged. If we could imagine anything akin — constantly — to the trade and 1 84 OTHER SUNS THAN OURS. counter-trade wind-zones on the earth in these belts, we might admit the same cause in explanation of them, the existence, nan^ely, of atmospheric currents from the equatorial towards the polar regions and from the polar towards the equatorial regions. But no one who has ever seen these cloud-belts through a good telescope can admit such an explanation for an instant. Yet there is only one way in which cloud-belts in the direction of a planet's rotation can possibly be explained. They must be due to differ- ences in the rates of rotation, causing a rush of cloud- masses forwards where regions of slower rotation are entered, and backwards where the regions reached are of more rapid rotation than those left. If these differences in rotational rate are not due to different distances from the axis in different latitudes — and manifestly they are not so caused — they must be due to difference of distance from the axis at different levels. Thus, then, in the multitudinous and ever- varying cloud-belts of the giant planets we have evidence of the vast range of distance, from and towards the centre, over which the cloud-masses around these planets can travel. When they rise they lag behind in long trailing masses ; when they descend they rush forwards : in either case until frictional resistances cause them to attain the same rate of rotational progress as the surrounding masses. But further, there is an argument in the case of the giant planets akin to that which was deduced from Professor G. H. Darwin's reasoning in the case of the sun. It has been shown that the perturba- tions of the movements of the inner satellites both of Jupiter and Saturn are absolutely inconsistent with the belief that the apparent globe of either planet is really occupied by the planet's mass. It is rendered certain by these researches that the real globe of either planet is very much smaller than the •globe we see. How much smaller has not yet been TWO SUNLIKE PLANETS. 185 ascertained by this method, but it is certain that the difference of size must be enormous and akin to that suggested by the other lines of reasoning considered above. In passing I may consider a difficulty which, though it could not obviate the force of the evidence already adduced, deserves attention, even were it for no other reason than this, that nearly always the study of difficulties leads to the recognition of new truths. It follows necessarily, from the vastness of the distances intervening between the visible surface of the giant planets and the actual surface of their globes, that the rarity of the outer regions which we see must be enormous. Nay, we seem forced to recognize here something like what we recognized in the case of the sun — the absence of any continuous atmospheric pressure throughout the cloud-laden regions surrounding the giant planets : for, according to all the known laws of gaseous pressure, the den- sities attained even at depths of a hundred miles below a cloud region existing at such pressures as prevail within our own cloud strata would be enor- mous. But if we admit such exceeding tenuity in the cloud region forming the visible outer surface of the giant planets, it seems at a first view as though the edge of these planets' discs ought not to be sharply defined, but resemble rather a soft haze or mist. But this idea will be corrected if we consider the real state of things in a tenuous cloud region, such as we suppose to form the outer part of the visible surface of Jupiter or Saturn. At the distance of either planet a depth of fifty miles would appear in the most powerful telescope as the finest possible line. But a line of sight passing fifty miles below the outermost cloud-surface on Jupiter would traverse no less than 4,000 miles within that surface, that is, a range of 4,000 miles of cloud. A line of sight passing Sven but five miles below the l86 OTHER SUNS THAN OURS. outer surface would pass through i,3CX) miles of cloud-strewn space. Is it likely that, however thinly the clouds might be strewn along a range of i,ooo miles, a line of sight could actually pierce through, so as to reach the region beyond .■■ If a line of sight could not so pass through, then that is equivalent to saying that the planet up to that distance — less than four or five miles from its apparent edge — would appear as if absolutely opaque. Yet four miles would be far beyond the power of the largest telescope to appreciate at Jupiter's distance. There- fore, unless the clouds are so thinly strewn that the eye can pierce through a range of i,ooo, nay of fully 4,000, miles of them, it is certain that the outline of the disc must appear as sharp and continuous as though the planet were a solid globe. In the case of Saturn the argument is even stronger, for he is very much farther away than Jupiter, so that lOO miles at Saturn's mean distance from us would look no larger than fifty miles at Jupiter's. Now a line of sight passing loo miles below the visible surface of Saturn, would have to traverse a range of some 4,700 miles of cloud-strewn space. But how if it shall appear that, though usually the cloud-layers around Jupiter suffice to give to the edge of his disc a well-defined appearance, not readily distinguishable from that of a solid globe (except, of course, for the parallel belts which are so strongly suggestive of a cloud-surface), yet at times the outer parts of Jupiter's disc are transparent to such a degree that a line of sight through 20,000 miles of the cloud-strewn region can yet pass onwards to detect faintly illuminated matter beyond.' This has happened, in four cases at least, — and it need hardly perhaps be said that its occurrence even once would suffice to prove all that such observation, even though repeated a hundred times, could establish. Let us consider the evidence. TWO SUNLIKE PLANETS. 1 87 The four moons of Jupiter, in their movements around the planet, pass athwart his face in one direction when on the hither side, and in the opposite direction when beyond him. They are all the time illuminated by the sun's light — except, of course, when they are in the planet's shadow. The degree of this illumination depends in part on the nature of their surfaces, but chiefly on their distance from the sun. Supposing them to have the same reflective capacity as our moon, which is probably very near the truth, the actual lustre of their surfaces under the sun's illumipating power is equal to one twenty- seventh of the lustre of the surface of the full moon ; therefore it would naturally be much more difficult to see one of them through a cloud of given density than to see our moon. Now Mr. Todd, of Adelaide, Government Observer for South Australia, in re- sponse to a suggestion of Sir George Airy's, devoted special attention for many years to the movements of the satellites of Jupiter, timing them carefully as they entered on the planet's face, or passed off) or hid behind one side of Jupiter's disc, or reappeared on the other side. While engaged on such work, Mr. Todd has on four occasions seen a satellite of Jupiter when on the farther side of the planet, and so situated that, were the planet opaque to its very edge, the satellite would be just invisible, its unseen surface lying just inside (but touching, in the optical sense) the outer edge of the planet. On each occa- sion Mr. Todd's observation was confirmed by his assistant, nearly as skilful an observer as himself. The instrument employed was a fine eight-inch telescope by Cooke, of York. The circumstances were on each occasion most favourable for distinct vision — and in the singularly pure air of South Australia an 8-inch telescope, in good observing weather, would do better than a 12-inch or even a 1 5 -inch telescope in our own hazy air. r88 OTHER SUNS THAN OURS. Now it may be easily calculated that to see the whole disc of a globe 2,000 miles in diameter within the apparent outline of a globe 80,000 miles in diameter, the smaller being at the time on the farther side, the line of sight must pass (in order to reach the innermost edge of the smaller globe) through no less than 25,000 miles of the substance, whatever it may be, forming the outer part of the larger planet. A cloud-strewn region, then, we certainly have, since Mr. Todd and his assistant could not possibly look through 25,000 miles of solid matter. But, moreover, the clouds must be strewn with exceeding tenuity. What sort of continuous cloud, for instance, can we imagine, through ten miles of which our own moon, twenty-seven times better lit than the satellites of Jupiter, could possibly be seen .■' The air we breathe at the earth's surface suffices, even when at its clearest, to cut off an appreciable amount of sunlight when the sun's rays have to traverse but a few hundred miles of it. Probably 20,000 miles of such air would barely let the sun's light through, and certainly such a range of air would hide the sun from view if the air were not clearer than it is on an average summer's day with us in England. The observation of a faint star through the outer parts of what looks like Jupiter's globe seems even more striking. Mr. EUery, of Melbourne, Govern- ment Observer for Victoria, has witnessed this re- markable phenomenon. A star so faint as to be barely visible on the darkest and clearest night to ordinary eyesight, was occulted by Jupiter a few years ago, the passage of the planet over the star being visible only in southern latitudes. Mr. EUery, armed with a telescope four feet in diameter (a reflector), watched the progress of Jupiter towards the star, expecting that the star would disappear the moment the planet's outline seemed to reach it. TWO SUNLIKE PLANETS. 1 89 But to his surprise the star continued , visible for several minutes, not finally disappearing until the Hne of sight to the star passed 800 miles below the apparent surface of the planet, t^e range of the line of sight in that case carried it along a distance within the planet of some 16,600 miles. What chance would an astronomer have of seeing a sixth- magnitude star, no matter what the power of the telescope he employed, through a cloud- stratum eight miles in thickness .' Yet here a star was seen through a range of more than two thousand times that dis- tance through cloud-strewn space ! But there is clear evidence that, however sharply defined the outlines of Jupiter and Saturn appear they are of the thinest conceivable cloud-texture. The outline of Jupiter has been observed by Schroter and others to be at times irregular — portions looking flattened, as though parts of the outer surface were chipped off. The outline of Saturn is often so dis- torted that it has assumed what has been called the square-shouldered aspect — a peculiarity of appear- ance which, as it has been observed by the Herschels, Airy, the Bonds, Coolidge, and other well-known observers, we cannot reject as simply the result of carelessness in observation ; nor can any form of illusion serve to explain it. Jupiter has been seen with a satellite just entered on his disc, and a few minutes later the same satellite — although, had there been no change in Jupiter's apparent outline, it would have been farther on the planet's face — has been seen oif the planet's disc, as though it had changed its mind and gone back. This cannot possibly be explained except by assuming that the outline of the planet is really formed of thinly strewn clouds to a depth of many thousands of miles, and that at times, over wide areas — many millions of square miles at a time it must be — the clouds which had formed the apparent outline change from the form 190 OTHER SUNS THAN OURS. of visible cloud to that of the invisible vapour of water, so causing the apparent outline to shrink to layers lying far lower down. And lastly, the con- dition of the outer layers of Jupiter's apparent globe has been shown by the way in which the satellites, as they have advanced through the outskirts of the shadow of that globe, have been seen to wax and wane in lustre before finally disappearing. We have other evidence, however, showing the partially sunlike condition of Jupiter. It is certain that a portion of the light which comes from the planet is inherent. We might fairly infer this from the vast superiority of the giant planet's light over that of Mars — for it is easy to calculate how much light we should get from Jupiter if his surface were of the same reflective quality — the same "whiteness," to use ZoUner's expression* — as that of Mars ; and we find that, when due account is taken of his much greater distance, Jupiter ought to be rather less brilliant than Mars when the latter is nearest both to the earth and to the sun (as in the autumn of 1877). ^ut Mars is never half as bright as Jupiter. Still, although this .would indicate a difference of surface, and therefore probably of condition, it would not of itself suffice to show that the light of Jupiter is inherent ; for if the surface of Jupiter were of driven snow, or even of pure white clouds, he would send us more light than we actually receive from him, yet without possessing any inherent light. It is when we notice that large portions of his surface are by no means white, that we infer from the total amount of light we receive from him that he is partly self-luminous. In the case of Saturn we have the same kind of evidence. But Jupiter's satellites have put this matter beyond a peradventure. They cast shadows, which — at * Albedo is the term he employs to indicate the reflective capacity of a surface. TWO SUNLIKE PLANETS. I9I least in the case of the nearer sateUites — should appear as round black spots, if Jupiter has no in- herent lustre. And so the shadows usually appear. Occasionally a tint of very dark purple has been suspected, but most observers, looking at the shadow of a satellite as seen in a good telescope, would re- gard the spot as absolutely black. The effect of contrast against the bright surface of the planet might make a really brown spot look black, but still, so far as one can judge from appearances, the shadow-spot looks ordinarily as black as a shadow thrown on an absolutely non-luminous body ought to appear. But we are able to correct this impres- sion very effectually by observing the satellites themselves when in transit across the planet's face. Near the edge of the disc they seem scarcely differ- ent in lustre from the planet itself, and sometimes they are quite lost when in that position. But when they pass well on to the disc they nearly always look much darker than the planet. On some parts of Jupiter's face they look actually black — that is, they appear as dark as a shadow. When, as often happens, the shadow lies close beside the disc of the satellite itself, it becomes easy to make exact comparison between the apparent tints of shadow and satellite. Under these circum- stances, it has occasionally happened that the satellite has been found to be at least as dark as its own shadow — in one or two cases it has appeared even darker. Now a single case of this kind suffices to prove all that could be shown by a hundred such cases. When we see side by side two round sur- faces, one of which we know to be the surface of a satellite (a body similar in surface-contour, no doubt, to our own moon), the other a part of Jupiter's surface which no sunlight can reach, and find that these two portions of surface show precisely the same degree of darkness, it becomes certain that 192 OTHER SUNS THAN OURS. from the region cast into shadow there comes as much light as from the sun-illumined surface of the satellite. This may be but about the thirtieth part of the apparent luminosity of our moon's surface (which is allowing the satellite of Jupiter a darker surface), but still it is something. Off the disc of Jupiter the satellites, one and all, look bright enough. Hence, we have it proved that in some cases as much light comes from Jupiter's surface, not as reflected sunlight, but on account of the inherent luminosity of that surface, as we get from the surface of a satellite reflecting sunlight to about one-thirtieth the amount reflected from our moon's surface. Now, that any inherent light should be found in any part of Jupiter's surface, there must be intense heat. This would be the case even if the surface we see were the actual surface of the planet ; for we know that no kind of rock-surface glows with inherent light unless it is very hot. But when we remember that the surface we see is a surface of cloud, that the cloud masses form layers probably thousands of miles in depth, it is evident that for any inherent light to show through these clouds (which cannot themselves be luminous) there must be an immense amount of light coming from the planet's real surface, which therefore must be intensely hot. It would seem as though, were the outer envelope of clouds removed, and the planet within seen from some point of view such that no reflected sunlight would come from him, he would be found to shine with considerable lustre of his own — probably akin to the kind of light which we get from the blood-red and garnet- tinted suns forming the fourth of Secchi's four orders of stars (as classified by their spectra). We may pause for a moment to consider a point touched on, in passing, above. The satellites show much more nearly the same lustre as Jupiter's surface, when near the edge than when near the middle. TWO SUNLIKE PLANETS. 1 93 This, of course, shows that Jupiter's disc is darker near the edge. Yet it does not look so. On the contrary, it looks so much brighter to ordinary eyesight near the edge that the French astronomer Chacornac was actually led to devise an ingenious theory in explanation of the supposed superiority of lustre there. I remember well how astonished was an astronomical friend of mine, who had spoken of this theory with approval, when I asked him to test the amount of the supposed increase of lustre near the edge, assuring him that he would find decrease instead of increase. He observed Jupiter with care- fully graduated darkening glasses, and found^ as I. had predicted, that the edges disappeared first. But the lesson derivable from Chacornac's mistake — a mistake into which many observers of Jupiter have fallen — is worth careful study. We learn how apt we are to be deceived by mere contrast. Just as the satellites of Jupiter in transit appear sometimes black by contrast with the planet's bright surface, and the shadows black when in reality only dark, so the parts near the edge appear brighter than the rest of the disc because contrasted with the dark sky around, though in reality darker. The general lesson is to beware always lest false theories should be suggested by illusions of the sort. The particular lesson is that the parts of Jupiter near the edge are darker than the rest, and the interpretation is, I take it, that these parts are to some degree transparent — Hot always or often, scarcely ever, perhaps, so trans- parent as they must have been on those occasions when Mr. Todd saw a satellite or when Mr. Ellery saw a faint star, through thousands of miles of this star-strewn outer covering of Jupiter, but still always transparent enough to allow much sunlight to pass through, and so to look darker than the rest of the planet's surface, because not reflecting so much sunlight. But the falling off of lustre may very O 194 OTHER SUNS THAN OURS, probably be in part also due to the circumstance that none of the planet's inherent light can come from the parts near the edge. The great red spot on Jupiter (see next Essay), of which during the last few years we have heard so much, is perhaps the most sunlike feature of all. It had a surface of about 150,000,000 square miles — three-fourths of the entire surface of our earth. From that surface came a ruddy light, which gave clear evidence, under keen spectroscopic cross- examination by Dr. Henry Draper, of being in part inherent. The singularly regular shape of the spot — a perfect ellipse — showed that it was due to expan- sive action of vaporous matter encountering, but overcoming, the resistance of the vaporous atmosphere around the region occupied by the spot. The only way, I believe, in which the form and colour and persistence of this great spot can be explained satis- factorily is by a theory akin to that suggested in my article on Sun-spots. It would seem that from deep down below the layers of cloud which, until 1876, had covered the region occupied afterwards by the spot, mighty masses of compressed vapour were flung upwards with intense energy, making their way through those layers of cloud, and rolling the clouds away on all sides till an enormous area was cleared. Through the region thus cleared of cloud we could see the ruddy light from the glowing surface within. So long as the forces at work below drove the clouds away from this region, the spot remained, retaining alike its colour and its singularly regular form. This lasted for more than five years, after which, though the spot did not disappear, yet it lost its lustre, while the irregularity of its shape showed that the vapor- ous masses flung up from below were no longer able to drive away, as before, the cloud-masses which were endeavouring (so to speak) to return to the region from which they had been driven. TWO SUNLIKE PLANETS. I9S The existence of disturbances such as this, so widespread, so long lasting, and giving evidence of such intense heat in the planet, must assuredly suffice to dispose of the belief that Jupiter is a world like our own, and to prove that, though not actually a sun, his condition is more nearly akin to the sun's than to that of our own earth. And what is thus proved of Jupiter is proved also of his brother giant Saturn, seeing that all the evidence shows Saturn and Jupiter to be in nearly the same stage of plane- tary life, Saturn, if anything, being the younger and more sunlike of the two. CHAPTER XIII. THE GREAT RED SPOT ON JUPITER. Professor Young in his farewell address to the Philadelphia meeting of the American Association for the Advancement of Science spoke of the great red spot which has for many years been the most remarkable feature of the planet Jupiter as a mys- tery " probably hiding within itself the master-key to the constitution of the great orb of whose inmost nature it was an outward and most characteristic expression." Without altogether accepting' the view of the red spot thus in strangely mixed metaphor presented, I must most thoroughly express my agree- ment with the opinion underlying Professor Young's rhetoric. The great red spot on Jupiter is un- doubtedly the mo.st mysterious of all the phenomena which even the Prince of Planets has presented to the student of astronomy. A vast opening, about 150 millions of square miles in extent, lasting many years, undergoing changes of shape and of position most remarkable in character, this great red spot undoubtedly contradicts emphatically all the old- fashioned ideas respecting Jupiter : and it as certainly presents many perplexing questions for those to answer who have adopted the more modern ideas. Yet it has always seemed to me that the more remarkable a phenomenon is, the better is it worth studying, and the more likely is it to reward careful ig6 THE GREAT RED SPOT ON JUPITER. ig7 Study by truthful information. A perplexing prob- lem of this kind may be compared to a complicated lock, which will not open to any ordinary key ; but when a key which will open it has been found then may we feel well assured that that key is the right one ; whereas, when a commonplace phenomenon has been accounted for, we can have no more cer- tainty that our solution is right than we can feel respecting a key (one, perhaps, among a dozen) which will open a lock of commonplace construc- tion. Without claiming that as yet the correct solu- tion of the problem of the great red spot has been found, or even that it can be, let us proceed to examine the problem with a view to the determina- tion of at least some of the points which the true solution must interpret. It was in the year 1876 that the great red spot was first observed — by Professor Pritchett of Glas- gow, Missouri. But I have before me a picture drawn by Professor Mayer, of the Stevens Institute, Hoboken, in 1871, wherein the place afterwards occupied by the spot is marked by an oval ring about the same size and shape. Whether this was actually the first beginning of the disturbance, or merely a coincidence^ cannot now be very readily determined : but it is at least worth noting, even though it should be no more than a coincidence. When first observed the great spot was symmetri- cal and well defined in shape, and of a somewhat strong ruddy tint. It was about 150 millions of square miles in extent (as was also the space enclosed within the oval ring seen by Professor Mayer). The greater axis of the oval was nearly three times as long as the shorter, but part of the difference was due to foreshortening. From a study of several hundred pictures I am led to conclude that the greater axis of the spot was not more than 2 J times longer than the shorter axis. Observations made by the late 1 98 OTHER SUNS THAN OURS. Professor H. Draper with the spectroscope seem to suggest that the light of the ruddy spot was in part inherent ; but others question whether the evidence accepted by Professor Draper was altogether valid. The spot continued visible, with little change of form or colour, for about six years, after which time, though it remained visible, it lost its symmetry of form and its characteristic ruddy tint. It was half- veiled for a time (at least in appearance) by the ex- tension of a cloud-belt lying north of it, as though this cloud-belt lying at a higher level had spread farther and farther over the spot. At present the spot, or rather the traces of the spot, can still be seen ; but it no longer presents any of the features, except enormous extension, which made it so remark- able a feature of the planet from 1876 to 1882. It was noteworthy that compared with the equa- torial markings on Jupiter the great spot seemed to lag, as if the equatorial cloud-belt were whirled round in a shorter time than the side zone on which the spot was seen. The first point to be noticed, in this remarkable phenomenon, appears to me to be that which the eye first recognizes, the symmetrical shape which the spot presented. Of course the spot was less symmetri- cal when seen with high powers than when observed with a small telescope ; but the symmetry of shape was none the less remarkable that it belonged to the spot as a whole rather: than to the spot' when minutely examined and largely magnified. Of course symmetry of form implies, in such a case, uniformity in the action of the forces at work in determining form — in this case, uniformity in the action by which the spot was produced. The path along which a projectile travels is uniform, apart from atmospheric resistance, because a uniform force is at work on the missile from the beginning to the end of its career. The action on the projectile THE GREAT RED SPOT ON JUPITER. T99 is along lines always parallel and vertical ; conse- quently the symmetry of the path is related to the vertical : a vertical line divides the path into two portions perfectly resembling each other. Again, the course of a planet round the sun, or of a ball swung round the centre, is symmetrical, because of the uniformity of the forces directed towards the centre. In one case the path is elliptical, in the other the path is circular, but in each case the central nature of the forces at work on the moving body tend to make the path symmetrical with refer- ence not to a line but to a centre. Again, observe a whirlpool, a tornado, the shapes of the clouds seen around a volcanic crater during eruption, and even the rounded forms of summer clf^ds, and we see in each case how tendencies towards or from a centre result in giving uniformity of shape to the aggrega- tion of matter resulting from such tendencies. The existence of a shape centrally symmetrical, whether circular or elliptical, implies in every case the exist- ence of forces tending either from or towards the centre. I know of no exception to this rule in nature, though of course artificial productions may show symmetrical forms without giving evidence of central forces. We may assume, then, that whatever were the forces at work in forming and maintaining the great red spot on Jupiter, they were related in some way to the centre of the oval region affected by them. They may have produced motion from that centre, or motion towards it, or there may have been move- ments of both sorts : but assuredly central forces were at work in some way or ways, where the great red spot was formed. While the symmetry of the spot's shape forces on us this general conclusion, the greater length of the spot in one direction than in another possesses also a special significance. 200 OTHER SUNS THAN OURS. The spot was manifestly a surface feature, in this sense that the layer of clouds in which the spot ap- peared as a sort of opening was part of the visible surface of the planet. This was shown by the cir- cumstance that as the spot drew near to arid passed the edge of the planet its outline remained distinctly visible. Had the spot been due to some formation lying below the surface clouds of the planet the spot could not have been thus traced up to the planet's edge. Of course this need not prevent us from recognizing the true cause of the spot as exist- ing far below the surface-level ; but manifestly the cloud layer was laid open at its outer surface. Now this being so, it is clear that were the forces which formed the spot all at work at that same surface level, and all acting from or towards a centre, we should expect them all to act with about the same degree of force, and the spot to have therefore a circular shape, unless we can recognize some likelihood that in different latitudes on Jupiter different conditions would exist, or, in other words, unless we can recog- nize the existence of zones on the planet akin in some sense to the trade and counter-trade wind zones on the earth. But although the most characteristic feature of Jupiter is the existence, almost always (if not always), of parallel bands or zones of clouds, diverse in their light-reflecting qualities, these zones have no per- manent position like the trade zones on the earth. They vary almost capriciously in position. Some- times there are but four or five of them, at others there are ten or ■ twelve or even more. We cannot recognize any permanent difference, then, in the condition of the various latitudes on Jupiter, to ac- count for the' oval figure maintained so long (six years at least) by the great spot. Yet we may still, or rather we must obviously, associate the lengthening of the great opening in a THE GREAT RED SPOT ON JUPITER. 201 direction parallel to the cloud zones, with the forces to which the existence of those clouds — as such — is due. Now it has always seemed to me that as the trade wind theory, once complacently advanced to explain the parallel belts of Jupiter and Saturn, most manifestly fails, we are driven to another interpreta- tion of the cloud-belts which is very significant in regard to Jupiter's condition. The trade winds and counter-trade winds, and the zones named after them, owe their existence to the difference between the rota- tional velocities in tropical regions which lie farther from the earth's axis and in temperate and arctic regions which lie nearer to that axis. The cloud- belts of Jupiter and Saturn must also be due to dif- ferences of rotational velocity, — not however between places in different latitudes on those planets, but between regions at different elevations in the cloud- envelopes of Jupiter and Saturn. We seem forced to admit, seeing that the belts are real, and no other way of accounting for their existence seems open to us, that there must be great movements of ascent and descent, in the cloud-laden envelopes surround- ing the giant planets. Matter carried upwards, as columns of ascending vapours, or missiles ejected to enormous heights from Jovian volcanoes, passing as such matter does from regions near the centre, where the motion of rotation is slower, to regions higher up, where the motion of rotation is more rapid, lag behind and cause a trailing of cloud forms towards the west. On the contrary, matter descending, as torrents ol falling rain, or matter falling back after ejection, would rush forwards and cause the cloud forms to be extended towards the east. Granted a suffi- cient range in height, whether in ascent or descent, and the parallelism thus arising would be as marked as we actually find it in the cloud-belts of the giant planets. But here certain questions arise which we must 202 OTHER SUNS THAN OURS. dispose of before we consider in this light the length- ening of the great spot. Can we imagine that the cloud-laden envelopes surrounding the giant planets have the enormous depth which this explanation would assign to them } The depth essential for this interpretation must bear a measurable proportion to the diameter of the planet. Less than at least a thousand miles (only a fortieth of the planet's diameter) would certainly not suffice ; for obviously the rotational velocities at the top and bottom of a cloud region one thousand miles high on Jupiter would not differ by more than about one-fortieth, or about 2|- per cent., whereas the sharp parallelism of the belts indicates quite a considerable difference of velocities. Taking, however, even a depth of one thousand miles for an atmosphere which at its highest part bears clouds such as exist in our very highest cloud-bearing atmospheric strata, say ten miles above the sea-level, we find a very remarkable state of things at a depth of even but a hundred miles below the visible cloud-surface of Jupiter, unless we suppose laws connecting density and pressure to be very different on Jupiter from the laws recognized here, — a supposition which must not unnecessarily be intro- duced. For our air ten miles above the sea-level has a density equal to about one-eighth the density of the air we breathe, the density doubling for each 3^ miles (or thereabouts) of descent. Taking this den- sity as that existing at the outskirts of the visible cloud-envelope of Jupiter, we find that with his known (and well-measured) gravitating energy the density would double for each mile and a half of descent. But say that it doubles only once for two miles of descent. Then four miles below the visible surface of the planet the atmospheric density would be already half (instead of one-eighth) of our air's at the sea-level ; six miles below it would equal the density of our air ; eight miles below it would be THE GREAT RED SPOT ON JUPITER. 2O3 double ; and only ten miles below the visible surface of Jupiter the density of his air would be four times the density of the air we breathe. After that, in the next ninety miles of descent, taking us only one hundred miles from the surface, there would be forty- five doublings of pressure and density, making the density millions of millions of times greater than that of air, thousands of millions of times greater than that of water, and hundreds of millions of times greater than the density of any terrestrial element. This of course is altogether preposterous ; but it shows that in one way or another we have to admit the existence of conditions in Jupiter which are utterly different from any known on earth. Less simple, but not less decisive, is the mathema- tical evidence adduced by Professor Geo. H. Darwin (son of Charles Darwin), who has shown that the movements of Jupiter's satellites would be other than they are if the mass of Jupiter were distributed uni- formly, or with any approach to uniformity, through- out the globe we measure as Jupiter's. Either there must be great compression towards the centre of Jupiter's globe, or the outer parts of the region within the cloud-surface we measure must be of very small density, the real globe beginning thousands of miles inside that envelope. I pass over for the moment the powerful argument derivable from the behaviour of Jupiter's satellites. But I must say that, in my opinion, when observers of great skill, like the late Admiral Smyth, Sir Thomas Maclear, Professor Pearson, Mr. Todd of Adelaide, Mr. Ellery of Melbourne, and the assist- ants of these last-named observers, record observa- tions, such as the reappearance of a satellite after its transit across Jupiter's disc had already begun, and the visibility of a satellite when behind the planet and well within the disc, and the visibility of a faint star through the outer envelopes of Jupiter, it seems 204 OTHER SUNS THAN OURS. to me idle to advance optical-illusion interpretations such as would barely avail to explain such pheno- mena recorded by the merest beginners with the telescope. Thus Mr. Todd, Government Observer at Adelaide, who has had more experience than any man living in observing transits and occupations of Jupiter's satellites (having specially devoted himself to the work, in response to an appeal of Sir George Airy's), records that on four occasions he saw a satellite pass behind the well-defined edge of the planet, the form of the satellite continuing visible, without distortion, until at last the whole satellite was thus seen through the outer parts of the planet, and that on each occasion his assistant, a very cautious and well-practised observer, saw the same pheno- menon. Reply is made that possibly Jupiter was a little out of focus, or his outline for some other reason indistinct, and the satellite not really seen within it, or possibly the observers (both of them !) mistook a false image of the satellite, the result of wearied eye, for the satellite itself. Surely we may say that such an explanation is inconsistent with all reasonable probabilities. A mere beginner in observation may have the edge of the planet out of focus, and suppose the blurred extension so produced to represent the real dimensions of the planet. But Mr. Todd is no mere beginner ; he is an " old hand," and an old hand at this particular kind of work. His assistant, again, is no beginner, but a practised observer. In hazy weather, again, even a practised observer might form an unsatisfactory estimate of the position of Jupiter's edge (though he would by no means see a clearly- defined outline to the satellite) ; but the weather was not hazy ; the sky was exceptionally clear and still (so Mr. Todd told me when I had the pleasure of meeting him at Adelaide in 1880). The wearied- eye theory would be quite out of the question in the case of a single observer of any skill ; but when Mr. THE GREAT RED SPOT ON JUPITER. 205 Todd, seeing the outline of the satellite through the outskirts of the planet, called his assistant to take his place at the telescope, there was no wearied eye with a false image of the satellite on it, at work, but a fresh eye, which had not been looking at a satellite of Jupiter's for some time ; and when Mr. Todd resumed his place at the telescope his eye, too, was practically a fresh one. So with other recorded cases, where skilful and well-practised eyes have ob- served phenomena which can only be explained by recognizing great tenuity in the outer cloud-laden regions of Jupiter, and a great extension of his gaseous surroundings in depth. We seem to have travelled a long way from the great red spot, but in reality all that we have been inquiring into since we left the spot bears importantly on our interpretation of that remarkable pheno- menon. When we see so many independent lines of evi- dence all pointing to the conclusion that that state of things prevails to which the only valid explanation of the shape of the great red spot had already led us, all reasonable doubt seems removed. We may rest assured, I think, that the red spot really owed its .symmetry of form to the central nature of the forces at work in forming it, and its elongated shape to the circumstance that regions at very different dis- tances from the planet's centre took part in forming the spot, uprising matter being left lagging westwards and down-sinking matter being hurried forward east- wards, instead of travelling with uniform velocities from the centre of disturbance. But now, as soon as we thus recognize a region below the visible surface of the planet as taking part in the disturbance indicated by the great spot, it is a natural thought that possibly the origin of the whole disturbance was not only below the visible surface of Jupiter but in the real globe of the planet. Let us 206 OTHER SUNS THAN OURS. see whether this idea leads us to any results which seem to correspond with the phenomena actually prer sented by the great spot. If the origin of the disturbance were in the real globe of Jupiter, then it must be presumed that the original disturbance was due to the intense heat per- vading the whole frame of the planet and was ex- plosive in character. An, outburst of compressed vapour from some gigantic volcano on Jupiter, carry- ing upwards vast vaporous masses to regions of much diminished pressure, would be followed by the rapid rush outwards of the expanding vapour, and by the sweeping away of the cloud masses which before had covered the region of disturbance, over an immense area. This area would be circular in shape in the case of a non-rotating planet, or in the case of com- paratively shallow vaporous envelopes like those which surround our earth. But in the case of masses of vapour flung upwards from the real sur- face of a rapidly rotating planet like Jupiter, with sufficient energy to burst their way through cloud- layers thousands of miles above that surface, there would undoubtedly be a marked trailing off of the vaporous masses westwards. They would acquire a westerly motion sufficing to give the region- of dis- turbance measurable superiority of length. in an east- and-west direction. For the westerly motion would continue after the upflung vapours had reached their greatest height. As a result of this process of westerly lagging the western end of the spot might be expected to be not quite so symmetrical in form as the easterly, a pecu- liarity which was actually noticed. Moreover, as the whole spot, or rather the whole of the cloud-region containing the spot, would drift steadily westwards, the planet turning all the time rapidly eastwards (one rotation in less than ten hours), it follows that the spot would have a slightly longer rotation period than THE GREAT RED SPOT ON JUPITER. 207 the equatorial markings, — which also was actually observed. According to this interpretation, the great red spot on Jupiter would indicate the occurrence of a tre- mendous outburst at the planet's real surface, an out- burst compared with which the great earthquake at Krakatoa was as child's play compared with the labours of many giants. That the outbursts at its commencement was sudden may be well believed. Yet judging from the long continuance of the great spot and of the sequent disturbance, the eruptive action must have lasted a long time. Of course it does not necessarily follow that the disturbance which caused the great opening in the cloud-envelope lasted as long as the opening itself. It may well be that the movements by which a disturbed cloud-belt on Jupiter returns to its normal condition are sluggish compared with the fierce action by which the disturb- ance is brought about, at (or it may be below) the fiery surface of the planet itself. Still the gigantic elliptic ring seen as early as 1 87 1, followed by a gigantic elliptic opening which remained for six years, and that again by a disturbed condition which has already lasted nearly three years and may last much longer, — all this seems quite inconsistent with the idea that the eruptive action giving birth (if my interpre- tation is correct) to this long-lasting disturbance was itself of short duration. And after all, it would not be very surprising, when we consider the enormous scale on which Jupiter is constructed, the tremendous heat which must in all probability pervade his whole frame, and the corre- spondingly increased duration of all internal disturb- ances, if the analogues on Jupiter of volcanic out- bursts which on the earth (so much smaller and now relatively aged) last often for many weeks, should on Jupitei' last, occasionally, for several years. Jupiter, according to all reasonable probability, must be a 208 OTHER SUNS THAN OURS. very young planet. If his planetary career began at the same time as the earth's, he is certainly much younger than our earth ; but even if he began his career as a planet millions of years before the earth, even then he would be younger than the earth in de- velopment. For those millions of years would be as nothing compared with the vast excess of the dura- tion of Jupiter's life-stages over the duration of the corresponding life-stages of the earth. Regarding Jupiter as in a much more youthful stage of planetary life than the earth is now passing through, and re- membering that even when Jupiter has reached the same stage as our earth his eruptive energies will be much greater than the earth's now are, we may well believe that the explosions now taking place on Jupiter must be on an incomparably grander scale than the . mightiest volcanic disturbances on the earth. Applying to Jupiter the reasoning which was applied to the disturbance of Krakatoa in 1883, we might readily find that even a greater dis- turbance than the Great Red Spot indicated, tre- mendous and far-reaching though that disturbance was, could be explained as resulting from a Jovian volcanic outburst, vaster and fiercer than terrestrial outbursts, because Jupiter is at once a mightier and a much younger planet. Let us look into this matter a little more closely : and first, let us ask if anything akin to the diffi- culty thus recognized in the case of Jupiter (and also in that of Saturn) exists elsewhere. Now in the case of the sun we have an orb which is probably in large part gaseous. We certainly have, visibly, a gaseous region thousands of miles in depth, even estimating the depth only from the visible surface of luminous cloud which we call the photosphere. And in the sun's case the attraction of gravity on the atmospheric region thus recognized is ten or twelve tfmes greater than the attraction on THE GREAT RED SPOT ON JUPITER. 209 the atmosphere of Jupiter. Therefore we have in the sun's case a much greater difficulty than in the case of Jupiter or Saturn. It is true that the intense heat pervading the whole frame of the sun suggests a way of meeting the difficulty which does not at first sight seem available in dealing with the giant planets. The laws which connect density and pressure at ordinary temperatures and at ordinary pressures may pro- bably fail altogether where the temperatures are so high and the pressures so enormous as they must be throughout the whole frame of the sun. We may say, indeed, as I have elsewhere shown, respecting the outer parts of the sun we see, what Professor Young said of the usually unseen corona, that if the term atmosphere be understood as we under- stand it when speaking of our own air, the gaseous regions forming the parts of the sun next within the photosphere do not form an atmosphere at all. Here are his remarks in regard to the corona, each one of them being fully applicable to the gaseous envelopes within the visible surface of the sun : — " Granting for the moment that the corona is in part and largely composed of an envelope of ex- ceedingly rare gaseous matter around the sun, — then we may call it an atmosphere, because being gaseous and attached to a cosmical body, it bears to that body a relation analogous to that borne by our atmosphere to the earth itself So far the term is a proper one. But now further, and on the contrary, the term ' atmosphere' carries with it to most per- sons certain ideas as to the distribution of tempera- ture, density, &c., in its different parts, which are based on the fact that our terrestrial atmosphere is nearly quiescent and in static equilibrium under the force of gravity, with a temperature not more than two or three hundred degrees above the absolute zero, while the density of the portion accessible to P 2IO OTHER SUNS THAN OURS. human observation is very considerable. On the sun the conditions are immensely, and almost incon- ceivably, different, so that the term ' atmosphere ' becomes a very misleading one. There the equili- brium, so far as there is any, is dynamical, not statical, and the density, temperature, and condition of the gaseous substance is far more nearly that of the residual gas in a Crookes's vacuum tube through which an induction coil is sending electrical dis- charges ; so different from that of ordinary air that Crookes thought he had found a fourth state of matter, bearing some such relation to the gaseous state as the gaseous does to the liquid." That this is so in regard to the sun is shown at once if we remember that the great openings we call spots disclose solar regions lying certainly not less than io,000 miles below the sun's visible sur- face. Now the strength and breadth of the hydrogen lines seen in the spectrum of the sun's coloured flames show that the hydrogen present there is not indefinitely rarer than hydrogen at the pressure of the air we breathe. Putting the pressure at the sun's visible surface at the millionth part of the atmospheric pressure on earth at the sea-level, and noting that gravity at the sun's visible surface is twenty-seven times gravity at the earth's, we find that at a depth of two or three miles below the sun's apparent surface, atmospheric pressure would be the same as at our sea-level were the same gases present, and temperature the same there as here. For in about the eighth of a mile the pressure would double, so that in 2^ miles there would be twenty doublings of pressure, raising the density from the millionth part of our air's to somewhat more than equality with the density of our air (two doubled, that double doubled, and so on, to twenty doublings, giving 1,048,576). In the next 2^ miles the pres- sure would be increased more than a millionfold, — THE GREAT RED SPOT ON JUPITER. 2 1 1 always assuming the conditions to hold which we recognize in our own atmosphere. This would happen in five miles out of 10,000 miles of depth, known to be occupied by gaseous matter. Even taking into account the tremendous heat prevailing in the sun, and the existence of much lighter gases in his surroundings than exist in our own air, we cannot escape conclusions scarcely less preposterous and assuredly quite as inadmissible as we have thus reached. If the pressure and densitj' did not double in less than a mile, or than ten miles, or even a hundred, — which is altogether impossi- ble, — we should still have, within a range of io,ooo miles, io,ooo,- or 1,000, or 100 doublings (in these cases respectively) ; and consequently even with the least of these numbers there would be a density at the base of the 10,000 miles exceeding a billion billion times* the density of our own air. Undoubtedly it is not in the high temperature of gases near the sun, or not in this only, that the solution of the enigma lies. We have also to take into account the freedom of movement which exists throughout the gaseous envelopes of the sun, and the constant movements which are no doubt taking place within these envelopes. In some such way, I think, we must encounter the difficulty, kindred in character if not so great in degree, which exists in Jupiter's case. We must admit the existence of intense heat throughout the gaseous surroundings of Jupiter, though we need not * In the first twenty, doublings equality with our atmo- spheric pressure would be attained, in the next twenty the pressure would be a million times greater, in the next a billion times, in the fourth twenty doublings the pressure would be a, trillion times, and in the last twenty it would be raised to a quadrillion times the pressure at our sea-level. (I use the English system of numeration, according to which a million raised to the second power is a billion, to the third power a trillion, to the fourth a quadrillion, and so forth.) P 2 212 OTHER SUNS THAN OURS. imagine that they are as hot as the gaseous envelopes of the sun, or that their temperature even approaches solar teniperatures. We must admit great freedom of motion within these gaseous and vaporous regions around Jupiter. So may we at once escape the difficulty which Jupiter assuredly presents, and be led to the conclusion which we had already reached from another side, — viz., that Jupiter's outer portions to a depth of many hundreds of miles within his visible surface do not belong to his real globe, but are mainly formed of gaseous, vaporous, and cloud- like matter. From yet other directions the same result has been reached, as I pointed out in my " Other Worlds than Ours," many years before the great spot had appeared. No one now supposes that Jupiter is made of other materials than those which form the earth on which we live, nor does any one now sup- pose that Jupiter is a hollow planet, as Sir David Brewster insisted. Yet if we do not adopt one view or the other we cannot possibly explain the small mean density of Jupiter otherwise than by assuming that the globe we measure for Jupiter is very much larger than the planet itself. Jupiter is 1,250 times as large as the earth, but only 310 times as massive. This, alone, proves that the real globe of Jupiter lies far within the cloud-strewn surface we measure. With the enormous attraction residing in 310 times the earth's mass, a globe of the same materials as our earth would be considerably denser instead of less dense than the earth. Assigning to Jupiter a density only equal to the earth's, its diameter would be little more than 50,000 miles. Jupiter's diameter is fully 80,000 miles. The distance of the cloud- strewn surface we see, from the real surface of the planet, cannot then, it would seem, be less than 15,000 miles (the difference between 40,000 miles and 25,000 miles, the halves of the just-named diameters). THE GREAT RED SPOT ON JUPITER. 213 The telescopic aspect of Jupiter corresponds much better with this startling result than with the idea that he has an atmosphere in the least resembling our earth's. CHAPTER XIV. A DEAD WORLD. The ancients fell into strangely incorrect ideas about the heavenly bodies. They chose the most beautiful of all the planets, beautiful alike in symmetry of shape and delicacy of colouring, as the emblem of misery and gloom, regarding Saturday, the day sacred to that planet god, as one on which all work was un- fortunate. They took, on the contrary, the most dis- appointing and unsatisfactory of all the sun's family as a fortunate orb, the emblem of love ; and although, strange to say, the day devoted to this planet (Friday) also, was deemed unfortunate for beginning a great work, or starting on a long journey, that was only because the next day, devoted to the unlucky Saturn, compelled rest ; and it is naturally unlucky to begin a great work if in a few hours you will have to rest from it. In like manner the ancients looked at the full moon, and because she was pale, and seemed so " silently and with so wan a face " to climb the sky, they thought she was cold. " Ice-cold Dian " she seemed to them at the very time when her surface, as modern science shows, is hotter a good deal than boiling water. I say this of the full moon in perfect consciousness that an American physicist, Mr. Langley, with an instrument which he calls the bolometer, finds the full moon colder than ice. I reject the evidence of that too delicate heat-measurer, and prefer the well- A DEAD WORLD. 215 attested teachings of the trustworthy old instrument, the thermopile, whose work has been tested and measured again and again and never found wanting in correct- ness. With too delicate a balance, you do not always know what moves it ; a breath may make some light substance you are weighing seem twenty times as heavy, or as light, as it really is. And so, I suspect, it has been with Mr. Langley's unpleasantly named heat-measurer. Some unobserved change near at hand has made the bolometer tell of cooling, where it should have told of heating if it had really recorded the influence of the moon's beams. Once an astro- nomer who supposed his delicate heat-measurements were telling him of the heat of stars, found that in reality he had been carefully measuring the heat generated by friction as he turned his telescope towards the star. Mr. Langley was making, we may be well assured, a similar mistake. Of course he thoroughly believes in the results he has obtained ; so fully does he believe in them that, supposing the cold he has found in the full moon to result from the thinness of the lunar air, or the absence of any air on the moon, he adopted the belief that rock surfaces at a great height above the sea-level do not get warm under the sun's rays, as science asserts. It was on a lofty peak of the Rocky Mountains, report says, that he maintained this argument. " A priori reasoning," he said, " may seem to show that these rocks around us must become hot under the sun's rays, but science should trust more in i posteriori evidence, the argu- ment from observed facts," — here he sat down, but for a singularly short time, on one of the rocks to which he had referred, — "I— I stand to it," he is reported to have continued, with some appearance of irritation, " despite of arguments d, priori or — or otherwise, that the moon must be intensely cold when she is full ; for my bolometer says so, and my bolometer is never mistaken." 2l6 OTHER SUNS THAN OURS. Sir John Herschel had long since shown, by a pro- cess of simple reasoning, that at the time of lunar midday the moon's surface must become at least as hot as boiling water. The present Lord Rosse, using one of the fine telescopes which his father con- structed (not, as has been mistakenly alleged, the great Parsonstown reflector*), arrived at a result corresponding to that which any one acquainted with the laws of physics could have anticipated. He ingeniously separated the heat which the moon re- flects from the heat which she radiates — that is, which she gives out as a warm body will. He found that the surface of the moon at lunar mid- day is Joo degrees hotter than the same surface at lunar midnight. (I mean degrees Fahrenheit, of course ; for the general reader, however intimate he may be with the rules for converting Centigrade or Reaumur, prefers to have no occasion to apply them.) Dividing these equally — as is only fair — on either side of nothing, we have a range from 250 degrees below nought, or 282 degrees below freezing, to 250 degrees above nought, or 38 degrees above boiling ! We may get less cold, by dividing unequally ; but then we get so much the more heat, and that would be quite unnecessary ; or we may get less heat, but then we get so much the more cold, and 250 degrees below zero would be cold enough in all conscience. * " Along whose tube a tall man may walk'without stooping," it is the custom to add. Doubtless, for a tunnelling along which tall men may walk conveniently, the great Rosse tele- scope is the best in existence. But, regarded simply as an " instrument for observing the heavenly bodies," which, per- haps, is more nearly what it was meant for, the " mighty mirror of Parsonstown " is not so satisfactory. If Sir William Herschel's great 40rfeet telescope "bunched a star into a cocked hat," Lord Rosse's still larger instrument played worse pranks still with the planets. " Zey show me somedings," said a well- known German astronomer, pathetically describing his expe- rience at Parsonstown, " zey show me somedings, and zey say, ' He is Saturn ' : and I believe zem." A DEAD WORLD. 2\^ The stoutest among us would be killed by ten seconds of such cold, as surely as he would be killed by one second in boiling water. The moon, which passes through the whole range of this change in a fort- night, is assuredly not a desirable place for creatures so unfortunately sensitive as we are to changes of temperature. All this, however, is not new but old. Till Mr. Langley came and set many doubting with his dread- ful bolometer, no one imagined that there could be life on a planet undergoing the vicissitudes of tem- perature which Sir John Herschel had correctly in- dicated, and Lord Rosse had demonstrated. That the moon, whatever her past history may have been, must now pass monthly through amazing vicissitudes of heat and cold, is certain, let Mr. Langley's bolo- meter say what it may. It is with the moon's past history we are concerned at present, not with these effects of the sun's action on the moon's dead body. For dead the moon assuredly is, now. It is as clear that there are none of the characteristics essential for life— as water, air, and reasonable ranges of tem- perature — on the moon at present, as it is that her surface has in the past been the scene of tremendous disturbance. That dead body of hers, carefully ex- amined, with due regard to the evidence which our earth also can give about planetary existence, may tell us as much as a post-mortem examination might tell the keenly observant anatomist of the past life of a human being. What may have been the precise features of the various eras of lunar life we may no more be able to tell than the anatomist can tell what thoughts passed through the brain which he dissects with his scalpel. But the broad outline of human life we may trace as surely as that anatomist can follow the stages by which the body he dissects had reached its final condition before death invaded life's sanc- tuary. 21 8 OTHER SUNS THAN OURS. It is here that, as it seems to me, new thoughts are suggested by knowledge recently acquired. The argument from analogy has, I think, been somewhat too narrowly applied to the moon and other planets, when they have been compared with our earth. In assuming that each planet has its youth, its mid-life, its old age, and finally its death, astronomers have doubtless been right enough ; but I think it is by no means so clear that they have been right in assuming (tacitly, perhaps, but still con- fidently) that the various stages in the lifetime of one planet resemble the corresponding stages in the life- time of another. A dog has stages of life correspond- ing to those of a man ; but a puppy is not a baby or even like one, a young whelp is unlike a lad, a dog is not a human being, and even a dead dog presents no very marked features of resemblance to a defunct man. I propose to consider here some points in which, most probably, the moon's life-history must have been entirely unlike the life-history of our earth. The considerations I shall urge may be applied, it will be found, to other planets, as well to those which are larger than the earth as to those which, like the moon, are very much smaller. I begin with some of the simple considerations in- volved, such as those relating to size, surface, sub- stance, and so on. Every one knows that the earth contains eighty- one times as much matter as the moon. I might dwell on the consideration that in gathering together her larger mass the earth must have become very much warmer than the moon ever was, even when the moon was at her youngest and hottest. For the celestial bodies owe their heat to their own energies in gathering their mass together ; and the greater the gathering energy the greater the developed heat, as certainly as the stronger a blacksmith's arm the greater the effect of his blows. A DEAD WORLD. 219 It would seem even that we have evidence of a still greater deficiency of original heat in the circum- stance that the moon not only had less energy with which to gather together her substance, but that, ha,ving , gathered it together, she has packed it less closely than the earth. If the moon were as compact as the earth she should have only an eighty-first part of the earth's volume. As a matter of fact she has fully/a forty-ninth part. The earth put in a suitable balance (I cannot indicate any suitable place for setting it) would be found to weigh about five and a half times as much as a globe of water of the same size. But weighing the moon, three and a half globes of water as large as the moon would bring down the scale on their side. Starting thus in her career of life with much less heat than the earth, the moon would cool also much more quickly. I do not mean by this that she would give out more heat, moment by moment, than the earth did at the same stage of her life ; but that she would be cooling faster in the same sense that a cup- ful of hot water cools faster than a bowlful, and a spoonful faster than a cupful. This is easily seen if we compare the moon when red-hot with the earth also red-hot, neglecting the effect of the air of either body in keeping in the heat, as clothing keeps in the heat of thehuman body. We maybe surethat the con- sideration thus neglected does not affect the general result ; for, certainly the moon was not better clothed (atmospherically) than the earth, at any correspond- ing stages of their lives — but the reverse. Our red- hot earth, then, had eighty-one times as much heat at that red-hot time than the moon at that (other) time when she was red-hot. And the earth was giving out thirteen and a half times as much heat, moment by moment, as the moon ; for in that degree her, surface exceeds the moon's. Now, if a man has ;£^8 1,000, while another has but ;£■ 1,000, and expends 220 OTHER SUNS THAN OURS. daily ^^ 13 los., while his poorer friend can only afford to expend £1 daily, the property of the former will last six thousand days, while that of the latter will in one thousand be completely exhausted. The richer man's money would last six times as long as that of the poorer man. The earth's heat would, in the same way, last six times as long as the moon's, at each stage of the cooling process. In this way the moon would manifestly age very fast as compared with the earth. If we imagine the moon and the earth at the same stage of planet life six millions of years ago, then in a million years from that time, or five millions of years ago, the moon was where the earth is now. What will five millions of years do for us, or rather for our home .' But even that way of putting it is not quite strong enough. Those five millions of years in the moon's history would correspond to six times as long — or to thirty millions of years — in the history of the earth ! Our globe will show marked signs of advanced age by the end of that time, I venture to predict, in calm as- surance that, at any rate, I shall not be contradicted by the evidence of observed facts. She would then be as far advanced in planetary life as the moon. Like the well-known calculation about wine, made under (pretended) vinous influence for All the Year Round several years ago, this calculation may be modified, yet the result come out unchanged. If I remember rightly, that calculation began : " Let us suppose there are eighty casks, or it may be eight hundred, or, perhaps, eight thousand," and so on. We might have begun our calculation by saying. Imagine the moon and the earth at the same age six millions, or it may be sixty millions, or perhaps six hundred millions of years ago. It really does not matter. Take the longest period. Six hundred millions of years ago the earth and moon were in the same stage of planetary life. Then we find that five A DEAD WORLD. 221 hundred millions of years ago the moon had reached the stage now reached by the earth, and three thou- sand millions of years hence the earth will have reached the same condition as the moon. Is not this, the reader may ask, a very different result from the former ? On the contrary, it is precisely the same result ! For, by our present assumption, the rate at which either planet ages is only one-hundredth of the rate we had before assumed. Hence the three thousand millions of years in our result indicate an amount of aging equivalent only to that resulting in thirty mil- lions of years according to our former assumption. Since, then, we are quite certain of this, that the time when earth and moon were equally advanced in planetary life must be set millions of years ago,, we are at least certain also of this, that our earth will not be so old as the moon now is, she will not be so wretchedly decrepit (if not so utterly dead) a world until many millions of years have passed. So far as this reasoning is concerned, the moon might have passed through a life much like that of our earth. She might well have had a life-bearing period akin to that through which the earth is pass- ing at present. True, the various stages of her life would be very much shorter, and we can very well believe that, therefore, the various forms of animal life which have been developed on the earth would not have had the same chance of being properly developed on the moon. Or, considering the pro- gress of a single race — our own — we can very well imagine that a being like man on the moon would not have sufficient time to pass through all the stages by which man has passed from the arboreal, hairy, pointed-eared, and four-legged ancestry now assigned him, to the civilized man of to-day, invent- ing every year more perfect instruments for destroy- ing his fellows. The Lunarian, thus understood, may have been no better fighter than the man of the 222 OTHER SUNS THAN OURS. caves, or even than the more advanced fighteis among the anthropoid apes, our cousins. In other words, he may have been a perfectly contemptible creature, instead of a being of imperially murderous instincts. But now a consideration comes in which suggests the idea that at no time could the forms of animal and vegetable life on the moon have resembled those on the earth. We must apportion to the moon no more than her fair allowance of water, no more than her fair allowance of air. And when we have done this we find strong reason for thinking that, though that allowance of water and of air may have done very well for the Lunarians, it would not have done at all for us. Let us begin with the water. The moon would have had one eighty-first part of the quantity of water which formed our earth's share. So far, good. That seems altogether fair. But observe. Our earth, with eighty-one times as much water, had a surface only thirteen and a half times as large over which to distribute that water in seas. It needs not even the ghost of Cocker to show that the earth had six times as much water per square mile. That of itself must have sufficed to make a very marked difference be- tween the moon's condition then and our earth's con- dition now. Nor does it seem at all likely that at any stage, either earlier or later, the moon would have had a better chance of doing well in the uni- verse than she had then- — that is, at the time when she had reached the same stage of cooling which the earth has reached now. But the want, or the short allowance, of water was as nothing compared with the thin air the Lunarians, if there ever were any, had to breathe. Of course, as regards the quantity of air, the reasoning is the same precisely as for water. Only one-sixth part of the quantity of air which we have on this earth per square mile was (on the average) A DEAD WORLD. 223 above each square mile of the moon. On the earth this would be a most serious matter. For the density of air depends on the weight of the total quantity- above the surface; so that the density of the air would be reduced to one-sixth part if the quantity of air above each square mile were no greater here than it probably was on the moon, in the correspond- ing part of her planetary life. Now in the highest ascent above the sea-level which men have yet made — the celebrated balloon ascent by Messrs. Coxwell ind Glaisher — the height attained was within a very few feet of the height of Mount Everest, the highest known peak of the Himalayas. At that height the air was reduced to little more than one-fourth its density at the sea-level. Mr. Glaisher fainted, and, if he had been alone, it would have been all up with him, even though the balloon might eventually have come down. Mr. Coxwell remained conscious, how- ever. Nay, with creditable zeal for science, he even urged the fainting meteorologist to " make one other little observation, now — do" to which, however, Mr. Glaisher only responded by fainting dead away. Mr. Coxwell began, he tells us, to feel rather blue. Looking at his hands, he perceived that they were quite blue. They were also powerless, which was an even more serious matter ; for it was necessary that the valve-strings should be pulled, if the conscious and the unconscious aeronauts were to be saved. Mr. Cox veil was equal to the occasion. Seizing hold of the valve-string with his teeth, he drew it — feebly indeed, but still so that it worked — and the balloon began to descend. He had saved himself and his companion, but only as by the skin of his teeth. Certainly this proved that no man could live, even for a few minutes, in air of only one-sixth of the density of the air at our sea-level — for that would be but two-thirds as dense as the air whose rarity so nearly killed our aeronauts. 224 OTHER SUNS THAN OURS. Even this, however, would be as nothing compared with the tenuity of the lunar air, if we are right in supposing that, at the corresponding stage of her planetary life, the moon had the same allowance of air as compared with her mass that our earth has now. For, the smaller quantity of air would be drawn down with smaller force, gravity at the moon's surface being only one-sixth part of gravity at the surface of our earth. Instead of the lunar air having one-sixth, it would only have one thirty-sixth, of the density of our air at the sea-level. Air so thin would not only be unbreathable by creatures like ourselves ; it would not support any kind of life known to us on earth, except such life as there is (if life it can be called) in rotifers and other such creatures, which can not only live with the smallest possible allowance of air, but resist with apparently unimpaired cheerfulness the action of a roasting heat and a much more than freezing cold, and can neither be boiled nor baked, nor drowned nor desiccated, to death. Consider, again, some of the unpleasant results of such extreme rarity or tenuity of the air. It may seem rather a convenience than otherwise that the mer- curial barometer would be only five-sixths of an inch in height. But the water barometer, instead of being, as with us, about thirty-three feet in height^ would have a height of less than one foot. Now this in itself would not signify, but it would mean (and this would signify) that one foot would be the extreme height to which a suction pump would raise water. What a nuisance that would be ! especially, too, where, as we have seen, water would not be very plentiful, and wells would have to be dug deeper than on the earth to reach it. Then drinking would be much more difficult — which might, however, be as well, where water would be so hard to get at — for in drinking we exhaust the air on one side of the water A DEAD WORLD. 225 in our drinking-vessel (the air inside the mouth), and the air on the other side (outside the mouth) oblig- ingly presses the water into the mouth, where we want it to go. But the air outside would not do this with sufficient energy if its density were reduced to one-sixteenth ; so that, in order to drink, one would have to tip the drinking-vessel up till the water ran out into the mouth, which would be, to say the least, an inelegant and unseemly way of drinking. In passing we may notice that, were it not for the unpleasant deficiencies here mentioned, creatures much larger than any on the earth might exist on the moon. A Brobdingnag on the earth would be by no means the terrible monster imagined by the inventive Captain Gulliver. From what is now known about the relation between the strength and the size of muscles, a Brobdingnag ten times as tall (and also ten times as broad and as thick) as a man, would be one hundred times as strong ; but he would be one thousand times as heavy. Thus he would be ten times as heavy as he ought to be, and would be just about as active as a man among ourselves who, weighing 140 lbs. (10 stone), had his weight increased by overloading to 1,400 lbs. The extra 1,260 lbs., or rather more than half a ton, would certainly not conduce to activity, insomuch that, when Gulliver first saw a Brobdingnag, he might have been sure that the creature's show of giant size must be hollow, or else its weight so great that movement would be impossible. Now on the moon such a Brobdingnag would weigh only 233^ lbs., and, so far as power of movement is concerned, would be like a man of 140 lbs. weighted down with only 94! lbs. Even this would be an awkward extra weight. But a lunar man six times as tall as one of ourselves would be all right, for he would be thirty-six times as strong, and also thirty-six times as heavy, so that he would be just as active as one of us. He would be no such Q 226 OTHER SUNS THAN OURS. contemptible giant as Jack the killer of giants dealt with so easily. He would stride six yards as easily as a man on earth strides one yard ; he could leap over a height of twenty-four feet as easily as an active youth leaps over a four-feet stile. The work he could do as a lunar engineer or road-maker would be something stupendous. With as much ease as a man on earth can raise a block of stone six inches in length, height and breadth, our lunar man could raise a cubical block one yard in the side, for such a block which on the earth would be 2i6 times as heavy, would on the moon be but thirty-six times as heavy,' and the lunar man would be thirty-six times as strong. A lunar Great Pyramid, representing the same amount of work as the Great Pyramid of Egypt, would be 1,500 yards in the side and 970 yards high. It would remain in the dry, thin air of the moon for as many hundreds of thousands of years as our Great Pyramid has lasted thousands ; and as it would be quite easily discernible with our telescopes, even with those of moderate power, there might, after all, be nothing very stupendously absurd in old Gruithuysen's idea that some of the features on the moon were the work of former lunar inhabi- tants. But unfortunately these large lunar men would have wanted plenty of air and plenty of water, especially when at work on great lunar edifices ; and if our estimate of the moon's past condition is sound, there would have been but very little water for them, and still less air. Perhaps all this may seem very little worth con- sidering. Of such speculations there is no end, said Sir John Herschel — after indulging in them to his heart's content. And certainly it seems somewhat idle to discuss the ways of lunar men, when we have every reason to think that, in a world whose various stages of life lasted so short a time as the moon's, no such creature as man could possibly have been de- A DEAD WORLD. 227 veloped, even if there had been the requisite supply of air and water. Let us rather consider, therefore, whether what is actually seen on the moon may not find its explanation in the circumstances we have been examining. That the stages of the moon's life would be very much shorter than those of our earth's life, follows, as we have seen, from the consideration of her smaller mass. But the stages of her life would not only be shorter than those of the earth, they would be different in character, because of the different amount of air and water, and also because the lunar atmosphere, before it became air such as we have (only rarer), must have been very different in quality from our air in its old unbreathable state. It has been shown by geologists that the various salts found in the sea must 'have belonged to it from the beginning. The familiar explanation that they were washed into the sea by rivers is no explanation at all ; as a matter of fact the substances thus washed down by rivers came to be present in the solid crust by the drying up of former seas. We can form from the constitution of sea-water some idea of the horrible kind of atmosphere which our earth originally po.sse.ssed. So also can we from many of the substances which we find in the earth's crust. There were sulphurous acid, and sulphuretted hydro- gen (savouring of rotten eggs, though it could not have suggested their presence in days before as yet any chickens had appeared), carbonic acid, hydro- chloric acid, and other highly disagreeable vapours. We have only to imagine what would happen if our earth were warmed up again, to see what must have been her state before she cooled. Nay, as she is kind enough to warm herself up locally at times, in sufficient degree to emit the very vapours which must of yore have been permanently outside her crust, we can tell by actual observation what they Q 2 228 OTHER SUNS THAN OURS. were. As the temperature beneath the earth's crust rises, the following gases and vapours are successively poured forth : carbonic acid gas (which chemists now call carbon dioxide), sulphurous acid, sulphu- retted hydrogen, boracic acid and hydrochloric acid. How pleasant an abode our earth would have been in her youth (independently of her high temperature) for breathing animals, may be inferred from the state of things formerly existing in the Avernian Lake, across which no bird could fly with life. The showers falling from the hot air of those days would be by no means showers of pure water. Boracic acid in the liquid state may not sound very terrible, but boracic acid has played the very — a verjr important part, we would say — in modifying rock substances in volcanic districts, and when it fell in showers must have greatly altered the character of the earth's hot crust. Sulphuric acid might be as innocent as rose-water for anything that its name may perhaps imply to many ; but when we speak of fiery hot vitriol, every one begins to recognize a substance that would probably have produced somewhat more marked changes on the hot crust of the earth than a drizzle of ordinary rain on the fields and plains of the earth produces to-day. Ammonia and various compounds of carbon, nitrogen and hydrogen, must have been present in the old atmo- sphere of the earth, and in various degrees of dilution with water must have been very effective as denuding agents. Now it is easily seen that the various stages of the earth's vulcanian history must not only have been very different from those of the moon's, but that the records left in the crust must have been very diff"er- ently treated. For example, the formation and the active existence of great volcanic craters on the earth probably preceded the formation of great mountain ranges. (I am not here referring to th