CHAPTER III.

THE EXPLORER OF THE HEAVENS.

“A knowledge of the construction of the heavens,” Herschel wrote in 1811, “has always been the ultimate object of my observations.” The “Construction of the Heavens”! A phrase of profound and novel import, for the invention of which he was ridiculed by Brougham in the _Edinburgh Review_; yet expressing, as it had never been expressed before, the essential idea of sidereal astronomy. Speculation there had been as to the manner in which the stars were grouped together; but the touchstone of reality had yet to be applied to them. This unattempted, and all but impossible enterprise Herschel deliberately undertook. It presented itself spontaneously to his mind as worth the expenditure of a life’s labour; and he spared nothing in the disbursement. The hope of its accomplishment inspired his early exertions, carried him through innumerable difficulties, lent him audacity, fortified him in perseverance. For this,

“He left behind the painted buoy That tosses at the harbour’s mouth,”

and burst his way into an unnavigated ocean.

Herschel has had very few equals in his strength of controlled imagination. He held the balance, even to a nicety, between the real and the ideal. Meditation served in him to prescribe and guide experience; experience to ripen the fruit of meditation.

“We ought,” he wrote in 1785, “to avoid two opposite extremes. If we indulge a fanciful imagination, and build worlds of our own, we must not wonder at our going wide from the path of truth and nature. On the other hand, if we add observation to observation without attempting to draw, not only certain conclusions, but also conjectural views from them, we offend against the very end for which only observations ought to be made.”

This was consistently his method. If thought outran sight, he laboured earnestly that it should be overtaken by it: while sight, in turn, often took the initiative, and suggested thought. He was much more than a simple explorer. “Even at the telescope,” Professor Holden says, “his object was not discovery merely, but to know the inner constitution of the heavens.” He divined, at the same time that he observed.

The antique conception of the heavens as a hollow sphere upon which the celestial bodies are seen projected, survived then, and survives now, as a convenient fiction for practical purposes. But in the eighteenth century the fiction assumed to the great majority a sort of quasi-reality. Herschel made an exception in being vividly impressed with the _depth_ of space. How to sound that depth was the first problem that he attacked. As a preliminary to further operations, he sought to fix a unit of sidereal measurement. The distances from the earth to the stars were then altogether unknown. All that had been ascertained was that they must be very great. Instrumental refinements had not, in fact, been carried far enough to make the inquiry profitable. Herschel did not underrate its difficulty. He recognised that, in pursuing it, _one hundredth of a second of arc_ “became a quantity to be considered.” Justly arguing, however, that previous experiments on stellar parallax had been unsatisfactory and indecisive, he determined to try again.

He chose the “double star method,” invented by Galileo, but never, so far, effectually put to trial. The principle of it is perfectly simple, depending upon the perspective shifting to a spectator in motion, of objects at different distances from him. In order to apprehend it, one need only walk up and down before a lamp placed in the middle of a room, watching its apparent change of position relative to another lamp at the end of the same room. Just in the same way, a star observed from opposite sides of the earth’s orbit is sometimes found to alter its situation very slightly by comparison with another star close to it in the sky, but indefinitely remote from it in space. Half the small oscillation thus executed is called that star’s “annual parallax.” It represents the minute angle under which the radius of the terrestrial orbit would appear at the star’s actual distance. So vast, however, is the scale of the universe, that this tell-tale swing to and fro is, for the most part, imperceptible even with modern appliances, and was entirely inaccessible to Herschel’s observations. Yet they did not remain barren of results.

“As soon as I was fully convinced,” he wrote in 1781, “that in the investigation of parallax the method of double stars would have many advantages above any other, it became necessary to look out for proper stars. This introduced a new series of observations. I resolved to examine every star in the heavens with the utmost attention that I might fix my observations upon those that would best answer my end. The subject promises so rich a harvest that I cannot help inviting every lover of astronomy to join with me in observations that must inevitably lead to new discoveries. I took some pains to find out what double stars had been recorded by astronomers; but my situation permitted me not to consult extensive libraries, nor, indeed, was it very material; for as I intended to view the heavens myself, Nature, that great volume, appeared to me to contain the best catalogue.”

On January 10th, 1782, he presented to the Royal Society a catalogue of 269 double stars, of which 227 were of his own finding; and a second list of 434 followed in December, 1784. All were arranged in six classes, according to the distance apart of their components, ranging from one up to 120 seconds. The close couples he regarded as especially adapted for parallax-determinations; the wider ones might serve for criteria of stellar proper movements, or even of the sun’s transport through space. For the purpose of measuring the directions in which their members lay towards each other--technically called “position-angles”--and the intervals separating them, he invented two kinds of micrometers, and notes were added as to their relative brightness and colours. He was the first to observe the lovely contrasted or harmonised tints displayed by some of these objects.

Herschel’s double stars actually fulfilled none of the functions assigned to them. He was thus left without any definite unit of measurement for sidereal space; and he never succeeded in supplying the want. In 1814 he was “still engaged,” though vainly, “in ascertaining a scale whereby the extent of the universe, so far as it is possible for us to penetrate into space, may be fathomed.” He knew only that the distances of the stars nearest the earth could not be less, and might be a great deal more, than light-waves, propagated at the rate of 186,300 miles a second, would traverse in three or four years. Only the _manner_ of stellar arrangement, then, remained open to his zealous investigations.

The initial question presenting itself to an intelligent spectator of the nocturnal sky is: What relation does the dim galactic star-stream bear to the constellations amidst which it flows? And this question our interior position makes very difficult to be answered. We see the starry universe, it has been well said, “not in _plan_ but in _section_.” The problem is, from that section to determine the plan--to view the whole mentally as it would show visually from the outside. The general appearance to ourselves of the Milky Way leaves it uncertain whether it represents the projection upon the heavens of an immense stratum of equally scattered stars, or a ring-like accumulation, towards the middle of which our sun is situated. Herschel gave his preference, to begin with, to the former hypothesis, and then, with astonishing boldness and ingenuity, attempted to put it experimentally to the proof.

His method of “star-gauging” was described in 1784. It consisted in counting the stars visible in successive fields of his twenty-foot telescope, and computing the corresponding depths of space. Admitting an average regularity of distribution, this was easily done. If the stars did not really lie closer together in one region than in another, then the more of them there were to be seen along a given line of vision, the further the system could be inferred to extend in that particular direction. The ratio of its extension would also be given. It would vary with the cube-roots of the number of stars in each count.

Guided by this principle, Herschel ventured to lay down the boundaries of the stellar aggregation to which our sun belongs. So far as he “had yet gone round it,” in 1785, he perceived it to be “everywhere terminated, and in most places very narrowly too.” The differences, however, between his enumerations in various portions of the sky were enormous. In the Milky Way zone the stars presented themselves in shoals. He met fields--of just one quarter the area of the moon--containing nearly 600; so that, in fifteen minutes, 116,000 were estimated to have marched past his stationary telescope. Here, the calculated “length of his sounding-line” was nearly 500 times the distance of Sirius, his standard star. Towards the galactic poles, on the contrary, stars were comparatively scarce; and the transparent blackness of the sky showed that in those quarters the supply of stars was completely exhausted. At right angles to the Milky Way, then, the stellar system might be termed shallow, while in its plane, it stretched out on all sides to an inconceivable, though not to an illimitable extent. Its shape appeared, accordingly, to be that of a flat disc, of very irregular contour, and with a deep cleft matching the bifid section of the Milky Way between Cygnus and Scorpio.

Herschel regarded this conclusion only “as an example to illustrate the method.” Yet it was derived from the reckoning-up of 90,000 stars in 2,400 telescopic fields! Its validity rested on the assumption that stellar crowding indicated, not more stars in a given space, but more space stocked in the same proportion with stars. But his hope of thus getting a true mean result collapsed under the weight of his own observations. “It would not be difficult,” he stated in 1785, “to point out two or three hundred gathering clusters in our system.” The action of a “clustering power” drawing its component stars “into many separate allotments” grew continually clearer to him, and he admitted unreservedly in 1802 that the Milky Way “consists of stars very differently scattered from those immediately about us.”

In 1811, he expressly abandoned his original hypothesis. “I must freely confess,” he wrote, “that by continuing my sweeps of the heavens my opinion of the arrangement of the stars has undergone a gradual change. An equal scattering of the stars may be admitted in certain calculations; but when we examine the Milky Way, or closely compressed clusters, it must be given up.”

And in 1817: “Gauges, which on a supposition of an equality of scattering, were looked upon as gauges of distance, relate, in fact, more immediately to the scattering of the stars, of which they give valuable information.”

The “disc-theory” was then virtually withdrawn not many years after it had been propounded. “The subtlety of nature,” according to Bacon’s aphorism, “transcends the subtlety both of the intellect and of the senses.” Herschel very soon perceived the inadequacy of his colossal experiment; and he tranquilly acquiesced, not being among those who seek to entrench theory against evidence. He found that he had undervalued the complexity of the problem. Yet it remained before his mind to the end. The supreme object of his scientific life was to ascertain the laws of stellar distribution in cubical space, and he devoted to the subject the two concluding memoirs of the sixty-nine contributed by him to the “Philosophical Transactions.” He was in his eightieth year when he opened, with youthful freshness, a new phase of arduous investigation.

“The construction of the heavens,” he wrote in June, 1817, “can only be known when we have the situation of each body defined by its three dimensions. Of these three, the ordinary catalogues give but two, leaving the distance or profundity undetermined.” This element of “profundity” he went on to determine by the absolutely novel method of what may be called “photometric enumeration.”

He began by asserting what is self-evident--that faint stars are, “one with another,” more remote than bright ones; and he argued thence, reasonably enough, that the relative mean distances of the stars, taken order by order, might be inferred from their relative mean magnitudes. Next he pointed out that more space would be available for their accommodation in proportion to the cubes of their mean distances. Here lies the value of the method. It sets up, as Herschel said, “a standard of reference” with regard to stellar distribution. It makes it possible to compare actual stellar density, at a given mean distance, with a “certain properly modified equality of scattering.” By patiently calling over the roll of successive magnitudes, information may be obtained regarding over- and under-populated districts of space.

Herschel’s reasonings on the subject are perfectly valid, but for practical purposes far in advance of the time. Their application demanded a knowledge of stellar light-gradations, which, even now, has been only partially attained. His surprising anticipation of this mode of inquiry came, therefore, to nothing.

His device of “limiting apertures” was a simultaneous invention. It was designed as a measure of relative star-distances. Pointing two similar telescopes upon two unequal stars, he equalised them to the eye by stopping down the aperture of the instrument directed towards the brighter object. Assuming each to emit the same quantity of light, their respective distances would then be inversely as the diameters of the reflecting surfaces by which they were brought to the same level of apparent lustre. But the enormous real diversities of stellar size and brightness render this plan of action wholly illusory. Even for average estimates, proper motion is apparently a safer criterion of distance than magnitude.

Herschel employed the method of apertures with better success to ascertain the comparative extent of natural and telescopic vision. The boundary of the former was placed at “the twelfth order of distance.” Sirius, that is to say, removed to twelve times its actual remoteness, would be a barely discernible object to the naked eye. The same star carried seventy-five times further away still, could be seen as a faint light-speck with his twenty-foot telescope; and, transported 192 times beyond the visual limit, would make a similar appearance in the field of the forty-foot. These figures, multiplied by twelve, represented, in his expressive phrase, the “space-penetrating power” of his instruments. Their range extended respectively to 900 and 2,300 times the distance of his “standard star.” He estimated, moreover, that, through the agency of the larger, light might become sensible to the eye after a journey lasting nearly seven thousand years! So that, as he said, his telescopes penetrated both time and space.

His last observation of the Milky Way showed it to be in parts “fathomless,” even with the forty-foot. No sky-background could be seen, but only the dim glow of “star-dust.” This effect he attributed to the immeasurable extension, in those directions, of the stellar system. The serried orbs composing it, as they lay further and further from the eye, became at last separately indistinguishable. Herschel, as has been said, formulated no second theory of galactic structure after that of 1784–5 had been given up. What he thought on the subject, with ripened experience for his guide, can only be gathered piecemeal from his various writings. The general appearance of

That “broad and ample road, whose dust is gold, And pavement stars,”

he described as “that of a zone surrounding our situation in the solar system, in the shape of a succession of differently condensed patches of brightness, intermixed with others of a fainter tinge.” And he evidently considered this _seeming_ to be in fair accord with reality. The “patches of brightness” stood for genuine clusters, incipient, visibly forming, or formed. They are made up of stars not less lustrous, but much more closely collected than Sirius, Arcturus, or Capella. The smallness of galactic stars would thus be an effect of distance, while their crowding is a physical fact. The whole of these clusters are (on Herschel’s view) aggregated into an irregular, branching ring, distinct from, although bound together into one system with the brilliants of the constellations. “Our sun,” he emphatically affirmed in 1817, “with all the stars we can see with the eye, are deeply immersed in the Milky Way, and form a component part of it.”

He took leave of the subject which had engrossed so many of his thoughts in a paper read before the Royal Society, June 11th, 1818. In it he showed how the “equalising” principle could be applied to determine the relative distances of “globular and other clusters,” provided only that their component stars are of the rank of Sirius. It is improbable, however, that this condition is fulfilled. In open groups, such as the Pleiades, enormous suns are most likely connected with minute self-luminous bodies; but the stars compressed into “globular clusters” appear to be more uniform, and may, perhaps, be intermediate in magnitude. Yet here again, the only thing certain is the prevalence of endless variety. Celestial systems are not turned out by the dozen, like articles from a factory. Each differs from the rest in scale, in structure, in mechanism. Attempts to reduce all to any common standard must then prove futile. Disparities of distance are of course concerned in producing their varieties of aspect, coarse-looking “balls of stars” being, necessarily, on the whole, less remote than those of smoother texture. Finer graining, however, may also be due to a composition out of smaller and closer masses. The two causes concur, and the share of each in producing a certain effect cannot, in any individual case, be apportioned.

Herschel was indeed far too philosophical to adopt rigid lines of argument. His reasoning did not extend “so far as to exclude a real difference, not only in the size, but also in the number and arrangement of the stars in different clusters.” Nevertheless, the discussion founded upon it is no longer convincing. To modern astronomers it appears to travel quite wide of the mark. Its interest consists in the proof given by it that the problem of sidereal distances, the original incentive to Herschel’s reviews of the heavens, attracted his attention to the very end of his thinking life. Throughout his long career, the profundities of the universe haunted him. He sought, _per fas, per nefas_, trustworthy measures of the “third dimension” of celestial space. The object of his search was out of reach, and has not even now been fully attained; but the path it led him by was strewn with discoveries.

The nets spread in his “sweeps” brought in, besides double stars, plentiful takes of the filmy objects called “nebulæ.” He recognised with amazement their profusion in certain tracts of the sky; increased telescopic light-grasp never failed to render a further supply visible; the heavens teemed with them. He presented a catalogue of 1,000 to the Royal Society in 1786, a second equally comprehensive in 1789, and a supplementary list of 500 in 1802. Their natural history fascinated him. What they were, what they had been, and what they should come to, formed the subject of many of those ardent meditations which supplied motive power for his researches. He not only laid the foundation of nebular science, but carried the edifice to a considerable height, distinguishing the varieties of its objects, and classifying them according to their gradations of brightness. Some presented a most fantastic appearance.

“I have seen,” he wrote in 1784, “double and treble nebulæ variously arranged; large ones with small, seeming attendants; narrow, but much extended lucid nebulæ or bright dashes; some of the shape of a fan, resembling an electric brush, issuing from a lucid point; others of the cometic shape, with a seeming nucleus in the centre, or like cloudy stars surrounded with a nebulous atmosphere; a different sort, again, contained a nebulosity of the milky kind, like that wonderful, inexplicable phenomenon about Theta Orionis; while others shine with a fainter mottled kind of light which denotes their being resolvable into stars.”

He, “through the mystic dome,” discerned

“Regions of lucid matter taking form, Brushes of fire, hazy gleams, Clusters and beds of worlds, and bee-like swarms Of suns and starry streams.”

Annular and planetary nebulæ were _as such_, first described by him. “Among the curiosities of the heavens,” he announced in 1785, “should be placed a nebula that has a regular concentric dark spot in the middle, and is probably a ring of stars.” This was the famous annular nebula in Lyra, then a unique specimen, now the type of a class.

The planetary kind, so-called from their planet-like discs, were always more or less of an enigma to him. The vividness and uniformity of their light appeared to cut them off from true nebulæ; on mature consideration, he felt driven to suppose them “compressed star-groups.” “If it were not, perhaps, too hazardous,” he went on, “to pursue a former surmise of a renewal in what I figuratively called the laboratories of the universe, the stars forming these extraordinary nebulæ, by some decay or waste of nature, being no longer fit for their former purposes, and having their projectile forces, if any such they had, retarded in each other’s atmospheres, rush at last together, and either in succession, or by one general tremendous shock, unite into a new body. Perhaps the extraordinary and sudden blaze of a new star in Cassiopeia’s Chair, in 1572, might possibly be of such a nature.”

At that early stage of his inquiries, Herschel regarded all nebulæ indiscriminately as composed of genuine stars. It was almost inevitable that he should do so. For each gain in telescopic power had the effect of transferring no insignificant proportion of them from the nebular to the stellar order. There was no apparent reason for drawing a line anywhere. The inference seemed irresistible, that resolvability was simply a question of optical improvement. As Messier’s _nébuleuses sans étoiles_ had yielded to Herschel’s telescopes, so--it might fairly be anticipated--the “milky” streaks and patches seen by Herschel would curdle into stars under the compulsion of the still mightier instruments of the future. He was led on--to use his own expressions in 1791--“by almost imperceptible degrees from evident clusters, such as the Pleiades, to spots without a trace of stellar formation, the gradations being so well connected as to leave no doubt that all these phenomena were equally stellar.” They were what Lambert and Kant had supposed them to be--island-universes, vast congeries of suns, independently organised, and of galactic rank. They were, each and all, glorious systems, barely escaping total submergence in the illimitable ocean of space. Under the influence of these grandiose ideas, Herschel told Miss Burney, in 1786, that with his “large twenty-foot” he had “discovered 1,500 universes!” Fifteen hundred “whole sidereal systems, some of which might well outvie our Milky Way in grandeur.”

His contemplations of the heavens showed him everywhere traces of progress--of progress rising towards perfection, then sinking into decay, though with a sure prospect of renovation. He was thus led to arrange the nebulæ in a presumed order of development. The signs of interior condensation traceable in nearly all, he attributed to the persistent action of central forces. Condensation, then, gave evidence of age. Aggregated stars drew closer and closer together with time. So that scattered or branching formations were to be regarded as at an early stage of systemic existence; globular clusters, as representing universes still in the prime of life; while objects of the planetary kind were set down as “very aged, and drawing on towards a period of change, or dissolution.”

Our own nebula he characterised as “a very extensive, branching congeries of many millions of stars,” bearing upon it “fewer marks of profound antiquity than the rest.” Yet, in certain regions, he found “reason to believe that the stars are now drawing towards various secondary centres, and will in time separate into different clusters.” As an example of the ravages of time upon the galactic structure, he adverted to a black opening, four degrees wide, in the Zodiacal Scorpion, bordered on the west by an exceedingly compact cluster (Messier’s No. 80), possibly formed, he thought, of stars drawn from the adjacent vacancy. The chasm was to him one of the most impressive of celestial phenomena. His sister preserved an indelible recollection of hearing him, in the course of his observations, after a long, awful silence, exclaim, “Hier ist wahrhaftig ein Loch im Himmel!” (Here, truly, is a hole in the sky); and he recurred to its examination night after night and year after year, without ever clearing up, to his complete satisfaction, the mystery of its origin. The cluster significantly located at its edge was lit up in 1860 by the outburst of a temporary star.

This was not the sole instance noted by Herschel of the conjunction of a chasm with a cluster; and chasms and clusters alike told the same story of dilapidation. He foresaw, accordingly, as inevitable, the eventual “breaking-up” of the Milky Way into many small, but independent nebulæ. “The state into which the incessant action of the clustering power has brought it at present,” he wrote in 1814, “is a kind of chronometer that may be used to measure the time of its past and future existence; and although we do not know the rate of going of this mysterious chronometer, it is, nevertheless, certain that since the breaking up of the Milky Way affords a proof that it cannot last for ever, it equally bears witness that its past duration cannot be admitted to be infinite.”

Thus the idea of estimating the relative “ages” of celestial objects--of arranging them according to their progress in development, originated with Herschel in 1789. “This method of viewing the heavens,” he added, “seems to throw them into a new kind of light. They are now seen to resemble a luxuriant garden which contains the greatest variety of productions in different flourishing beds; and one advantage we may at least reap from it is that we can, as it were, extend the range of our experience to an immense duration. For, is it not almost the same thing whether we live successively to witness the germination, blooming, foliage, fecundity, fading, withering, and corruption of a plant, or whether a vast number of specimens, selected from every stage through which the plant passes in the course of its existence, be brought at once to our view?”

But while he followed the line of continuity thus vividly traced, another crossing, and more or less interfering with it, opened out before him. The discovery of a star in Taurus, “surrounded with a faintly luminous atmosphere,” led him, in 1791, to revise his previous opinions regarding the nature of nebulæ. He was not at all ashamed of this fresh start. No fear of “committing himself” deterred him from imparting the thoughts that accompanied his multudinous observations. He felt committed to nothing but truth. He was advancing into an untrodden country. At every step he came upon unexpected points of view. The bugbear of inconsistency could not prevent him from taking advantage of each in turn to gain a wider prospect.

Until 1791 Herschel never doubted that gradations of distance fully accounted for gradations of nebular resolvability. He had been led on, he explained, by almost imperceptible degrees from evident clusters to spots without a trace of stellar formation, no break anywhere suggesting the possibility of a radical difference of constitution. “When I pursued these researches,” he went on, “I was in the situation of a natural philosopher who follows the various species of animals and insects from the height of their perfection down to the lowest ebb of life; when, arriving at the vegetable kingdom, he can scarcely point out to us the precise boundary where the animal ceases and the plant begins; and may even go so far as to suspect them not to be essentially different. But, recollecting himself, he compares, for instance, one of the human species to a tree, and all doubt upon the subject vanishes. In the same manner we pass by gentle steps from a coarse cluster to an object such as the nebula in Orion, where we are still inclined to remain in the once adopted idea of stars exceedingly remote and inconceivably crowded, as being the occasion of that remarkable appearance. It seems, therefore, to require a more dissimilar object to set us right again. A glance like that of the naturalist, who casts his eye from the perfect animal to the perfect vegetable, is wanting to remove the veil from the mind of the astronomer. The object I have mentioned above is the phenomenon that was wanting. View, for instance, the nineteenth cluster of my sixth class, and afterwards cast your eye on this cloudy star, and the result will be no less decisive than that of the naturalist. Our judgment, I venture to say, will be that _the nebulosity about the star is not of a starry nature_.”

In this manner he inferred the existence of real nebulous matter--of a “shining fluid” of unknown and unimaginable properties. Was it perhaps, he asked himself, a display of electrical illumination, like the aurora borealis, or did it rather resemble the “magnificent cone of the zodiacal light?” A boundless field of speculation was thrown open. “These nebulous stars,” he added, “may serve as a clue to unravel other mysterious phenomena.”

As their close allies, he now recognised planetary nebulæ, the “milkiness, or soft tint of their light,” agreeing much better with the supposition of a fluid, than of a stellar condition. And he rightly placed in the same category the Orion nebula, and certain “diffused nebulosities” which he had observed just to tarnish the sky over wide areas. These last might, he considered, be quite near the earth, and the object in Orion not more distant than perhaps an average second magnitude star.

The relations of the sidereal to the nebular “principle” exercised Herschel’s thoughts during many years. He had no sooner reasoned out the existence in interstellar space of a rarefied, self-luminous substance, than he began to interrogate himself as to its probable function. Nature was to him the expression of Supreme Reason. He could only conceive of her doings as directed towards an intelligible end. Hence his confidence that rational investigation must lead to truth.

Already in 1791 he hinted at the conclusion which he foresaw. The envelope of a “cloudy star” was, he declared, “more fit to produce a star by its condensation than to depend upon the star for its existence.” And the surmise was confirmed by his detection, in a planetary nebula, of a sharp nucleus, or “generating star,” possibly to be completed in time by the further accumulation of luminous matter.

His conjectures developed in 1811 into a formal theory. The cosmical fluid was met with in all stages of condensation. Nebulous tracts of almost evanescent lustre were connected in an unbroken series with slightly “burred” objects, wanting only a few last touches to make them finished stars. The extremes, as he said, had been, by his “critical examination of the nebulous system,” “connected by such nearly allied intermediate steps, as will make it highly probable that every succeeding state of the nebulous matter is the result of the action of gravitation upon it while in a foregoing one.”

In 1814 he traced the progress towards maturity of binary systems. Originating in double nebulæ incompletely dissevered--Siamese-twin objects, of which he had collected 139 examples--they next appeared as nebulously-connected stars, finally as a pair materially isolated, and linked together by the sole tie of gravitation. Scattered clusters represented, in his scheme of celestial progress, a state antecedent to that of globular clusters. “The still remaining irregularity of their arrangement,” he said, “additionally proves that the action of the clustering power has not been exerted long enough to produce a more artificial construction.” He made, too, the important admission that clusters apparently “in, or very near the Milky Way,” were truly part and parcel of that complex agglomeration.

But what of his “fifteen hundred universes,” which had now logically ceased to exist? The stellar and nebular “principles” had virtually coalesced; both were included in the galactic system. The question of “island universes” was accordingly left in abeyance; although Herschel certainly believed in 1818 that among the multitude of “ambiguous objects”--we should call them irresolvable nebulæ--many exterior firmaments were included. Yet what he had ascertained about the distribution of nebulæ should alone have sufficed to shatter this remnant of a conviction.

The fact became clear to him during the progress of his “sweeps” that nebulæ, to some extent, _replace stars_. He found them to occur in “parcels,” more or less embedded with stars, “beds” and “parcels” together being surrounded by blank spaces. This arrangement grew so familiar to him that he used to notify his assistant, when stars thinned out in the zone he was traversing, “to prepare for nebulæ.” A wider relationship, brought within view by the large scale of his labours, was defined by his fortunate habit of charting, for convenience of identification, each newly-discovered batch of nebulæ.

“A very remarkable circumstance,” he wrote in 1784, “attending the nebulæ and clusters of stars, is that they are arranged into strata, which seem to run on to a great length; and some of them I have already been able to pursue, so as to guess pretty well at their form and direction. It is probable enough that they may surround the whole apparent sphere of the heavens, not unlike the Milky Way.”

In the following year he spoke no longer of a zone, but of two vast groupings of nebulæ about the opposite poles of the Milky Way. That is to say, where stars are scarcest nebulæ are most abundant. The correspondence did not escape him; but he did not recognise its architectonic meaning. He had traced out the main plan of the stellar world; he had discovered, not merely thousands of nebulæ, but the nebular system; he had shown that stars and nebulæ were intimately associated; he had even made it clear that nebular distribution was governed by the lines of galactic structure. It only remained to draw the obvious inference that these related parts made up one whole--that no more than a single universe is laid open to human contemplation. This was done by Whewell thirty years after his death.