heavens consist of spherical or elliptical forms presenting every variety of elongation and central condensation. Of these a great number have been resolved into stars, and a great many present that mottled appearance which renders it certain that an increase of optical power would decompose them. Those which resist do

so on account of the smallness and closeness of the stars of which they consist.

Elliptical nebulæ are very common; by much the finest may be seen near the star v in the girdle of Andromeda. It is visible to the naked eye, and has frequently been taken for a comet. With a low optical power it has the spindle-shaped form of fig. 6, plate 5, the brightness being at first gradually and then rapidly condensed towards the centre, so that it has been compared to a star shining through horn, but had never appeared resolvable even with high optical powers till Mr. Bond examined it at the observatory of Cambridge in the United States. He found that its brightness extends over 23 degrees in length, and more than a degree in breadth, including two small adjacent nebulæ ; so that it is oval. It is strongly and rapidly condensed into a nucleus on its northern side; and although it was not all resolved, it was seen to be strewed over with star dust, or extremely minute visible stars, which leaves not a doubt of its being a starry system. The most remarkable part of Mr. Bond's discovery are two very narrow dark lines which extend along one side of the oval parallel to its major axis. These black streaks, difficult to distinguish, indicate a stratified structure, and are not the only instance of that arrangement in nebulæ. Fig. 1, in plate 9, is from Mr. Bond's drawing of this nebula.

Multitudes of nebulæ appear to the unassisted eye, or are seen with ordinary telescopes, like round comets without tails; but when viewed with powerful instruments they convey the idea of a globular space, insulated in the heavens and full of stars, constituting a family or society apart from the rest, subject only to its own internal laws. To attempt to count the stars in one of these globular clusters, Sir John Herschel says, would be a vain task; they are not to be reckoned by hundreds. On a rough computation, it appears that many clusters of this description must contain ten or twenty thousand stars compacted and wedged together in a round space, whose apparent area is not more than a tenth part of that covered by the moon; so that its centre,

where the stars are seen projected on each other, is one blaze of light. If each of these stars be a sun, and if they be separated by intervals equal to that which separates our sun from the nearest fixed star, the distance which renders the whole cluster barely visible to the naked eye must be so great, that the existence of such a splendid assemblage can only be known to us by light which must have left it at least a thousand years ago. These magnificent globular or spheroidal aggregates of stars are so arranged that the interior strata are more crowded and become more nearly spherical as they approach the centre. A most splendid object of this nature may be seen in the constellation Hercules (N. 235).

Of 131 of these magnificent objects in the southern hemisphere, two of them are pre-eminently splendid. The globular cluster of a Centauri is beyond comparison the finest of its kind: it is perfectly spherical, and occupies a quarter of a square degree; the stars in it are literally innumerable, crowding and densely aggregated towards the centre; and, as its light is not more to the naked eye than that of a star of the 4th or 5th magnitude, their minuteness is extreme. It has a dark hole in its centre, with a bridge of stars across,—a circumstance peculiar to this cluster.

Lacaille's globular cluster, or 47 Toucani, is completely insulated in a very dark part of the sky not far from the lesser of the Magellanic clouds. The stars, which are of the 14th magnitude, immensely numerous, compressed and white, form three distinct stages round a centre, where they suddenly change in hue, and form a blaze of rose-coloured light. One cluster consists of large ruddy stars and small white ones; another of greater beauty consists of shells or coats of stars of the 11th and 15th magnitude. There are thirty globular clusters of extreme beauty collected within a circular space of not more than eighteen degrees radius, which lies in the part of the sky occupied by the constellations Corona Australis, the body and head of Sagittarius, the tail of Scorpio, part of Telescopium and Ara. The Milky Way passes diametrically across the circular area in question, which gives prodigious brilliancy to this part of the sky. For besides these globular clusters, which all lie in the starry part, and not in the dark spaces, there are the only two annular nebulæ known to exist in the southern hemisphere. No part of the heavens is fuller of objects beautiful and remarkable in themselves, and ren

dered still more so by their mode of association, and by the peculiar features assumed by the Milky Way, which are without a parallel for richness and magnificence in any other part of the sky. Some of the globular clusters are so remote that the stars are scarcely discernible-mere star dust. There is a double globular cluster in the southern hemisphere of very small dimensions, separated by a minute interval,- -a combination which suggests the idea of a globular cluster revolving about a very oblate spheroidal one in the plane of the equator, and in an orbit which is circular, and seen obliquely like the central nebula itself, with a diameter somewhat more than four times the latter,-a stupendous system doubtless, but of which the reality can hardly be supposed improbable.

There appears to be some connexion between ellipticity of form and difficulty of resolution, for spherical clusters are in general easily resolved into their component stars, while there is scarcely an instance of an elliptical cluster yielding except to a very high optical power. Vast masses of the nebulæ have never been resolved. Lord Rosse's great telescope has resolved parts of the nebula of Orion, and various others which had not yielded to instruments of less power; it enables the astronomer to penetrate farther into space, and shows objects with greater clearness, than any other. But, excellent as this instrument is, thousands of nebulæ are not to be resolved even by it. Those who imagine that any work of man can resolve all the nebulous matter in the heavens must have a very limited idea of the extent and sublimity of creation.

Innumerable nebulæ in both hemispheres take the form of clusters of stars, but are totally different from the globular clusters, inasmuch as they are of irregular form and follow no uniform law of condensation. The Pleiades is an instance in our own stellar system; for although only 7 or 8 stars are visible to the naked eye, telescopes show that more than 200 belong to the group. In the constellation Cancer there is a luminous spot called the Præsepe or Beehive, which a very low power resolves into stars; and the constellation Coma Berenices is another stellar group. Many are of exquisite beauty, as that round a Crucis, which, though consisting of only a hundred and ten stars, is like a piece of fancy jewellery, from the colours of the stars, which are greenish white, green, blueish green, and red. Many of these

clusters contain thousands of stars, and are frequently in the poorer parts of the sky, as if in the course of ages the stars had been attracted towards a centre.

The existence of every degree of ellipticity in the nebula-from long lenticular rays to the exact circular form—and of every shade of central condensation, from the slightest increase of density to apparently a solid nucleus-may be accounted for by supposing the general constitutions of those nebula to be that of oblate spheroidal masses of every degree of flatness from the sphere to the disc, and of every variety in their density and ellipticity towards the centre. It would be erroneous, however, to imagine that the forms of these systems are maintained by forces identical with those already described, which determine the form of a fluid mass in rotation; because, if the nebulæ be only clusters of separate stars, as in the greater number of cases there is every reason to believe them to be, no pressure can be propagated through them. Consequently, since no general rotation of such a system as one mass can be supposed, it may be conceived to be a quiescent form, comprising within its limits an indefinite number of stars, each of which may be moving in an orbit about the common centre of the whole, in virtue of a law of internal gravitation resulting from the compound gravitation of all its parts. Sir John Herschel has proved that the existence of such a system is not inconsistent with the law of gravitation under certain conditions.

The distribution of the nebulæ over the heavens is even more irregular than that of the stars. In some places they are so crowded together as scarcely to allow one to pass through the field of the telescope before another appears, while in other parts hours elapse without a single nebula occurring. They are in general only to be seen with the best telescopes, and are most abundant in a zone whose general direction is not far from the hour circles Oh and 12h, and which crosses the Milky Way nearly at right angles. Where that nebulous zone passes over the constellations Virgo, Coma Berenices, and the Great Bear, they are to be found in multitudes.

The nebulous system is nearly divided into two parts by the Milky Way. One-third of the whole visible nebulous contents of the heavens forms a broad irregular mass, interspersed with vacant intervals, which fills about an eighth of the surface of the

northern hemisphere. It occupies the constellations Leo, Leo Minor, the body, tail, and hind legs of Ursa Major, the nose of Camelopard, the point of the tail of Draco, Canis Venatica, Coma Berenices, the preceding leg of Boötes, and the head, wings, and shoulder of Virgo, which is the richest part. There is a lesser nebulous region in this hemisphere, but entirely separated from the preceding, which occupies the chest and wing of Pegasus, the constellations Pisces and Andromeda. If we could imagine the ring or zone of the Milky Way to encircle or coincide with the horizon, the great nebulous mass would form a canopy over head, descending down to a considerable distance on all sides, chiefly towards the north pole; and the richest part, which is in Virgo, would then be directly over head in the north pole of the Milky Way, that is in 12h 47m right ascension, and 64° north polar distance.

With the exception of the Magellanic clouds, there is a much greater uniformity in the distribution of the nebula in the southern hemisphere than in the northern. They are separated by spaces of vacuity of greater or less dimensions. One of these barren regions extends for nearly fifteen degrees all around the south pole, and close on its border; the lesser of the Magellanic clouds occurs completely insulated; while the greater Magellanic cloud is in connexion with something approaching to a zone of connected nebulous patches which extends along the back of Doradus, through a portion of Horologium and Eridanus, part of Fornix, and over the paws of Cetus to the equator, where it unites with the nebulous regions of Pisces.

The Magellanic clouds form two of the most striking features in the southern hemisphere; both of these nebulæ are visible to the unassisted eye, being nearly of the same intensity as the brighter portions of the Milky Way; but the smaller is entirely effaced by moonlight, and the larger nearly so. They are altogether unconnected with the Milky Way and with one another. The Nubecula Major is far superior to the Nubecula Minor in every respect, though they are similar in internal structure. The former consists of large tracts and ill-defined patches of irresolvable nebulæ, and nebulosity in every stage of resolution, up to perfectly resolved stars like the Milky Way; and also of regular and irregular nebulæ, properly so called; of globular clusters in every stage of resolvability; and of clustering groups suffi