bling that of the solid cloven disc exhibited in Figs. 4 and 5. The latter figure, being a side view, gives the figure of the

Fig. 4.

Fig. 5.

imaginary section which appears so often in works on astronomy. In Figs. 3 and 5 the small circle near the centre indicates the probable extent (on the theory in question) of the sphere within which stars down to the fourth magnitude may be supposed to be included.

The main difficulties in attempting to form an estimate of the real configuration of the galactic nebula are those which have been already mentioned. Have we evidence confirming or disproving (1) the tendency to clustering suggested by the elder Herschel, (2) the possible variability among star-magnitudes, and (3) of influences exerted by large stars in guiding or swaying others? It appears to me that there are indications of a very obvious and important character, which have been either altogether unnoticed, or much less noticed than they deserve. In considering these indications, I would refer the reader chiefly to Figs. 1 and 2; but some portions of the evidence cannot be thoroughly understood without reference to star-maps, in which (at a single view, if possible) the course of the Milky Way is exhibited in a manner which enables us at once to determine its relations to the constellations not included in these figures. I write with my black star-maps before me, my object being to consider the special evidence afforded by stars of the leading magnitudes. It may seem (and, indeed, on the assumption of any approach to uniformity in the true magnitudes, or distribution of stars, it must necessarily be) a very imperfect method to refer to star-maps including only the first five magnitudes, still more to consider the first four magnitudes which are alone represented in Fig. 1. It is

obvious that if our sun is placed within such a stratum as is exhibited in Figs. 3, 4, and 5, no evidence whatever as to the structure of that stratum can be afforded by considering the comparatively few stars included within the small central circle. But it is this very fact on which I wish to dwell. If any connection does appear between the configuration of our galaxy, and the arrangement of stars which are assumed to be much nearer to us than the Milky Way, it will be obvious that we must somewhat modify the views illustrated by these figures.

Now, taking Figs. 1 and 2, I think one can trace a connection between the stars there depicted, and that stream of nebulous light which the view we are examining teaches us to consider as at an indefinite distance beyond those stars. In the northern portion, perhaps, the connection is not very remarkable. We see that a large number of the brighter stars lie on or near the Milky Way, but it would require the examination of a somewhat wider zone than that here presented to exhibit this arrangement as positive evidence of aggregation. However, I think no one who has attentively examined the glories of Orion, the richly-jewelled Taurus, the singular festoon of stars in Perseus, and the closely-set stars of Cassiopeia, but must have felt that the association of splendour along this streak of the heavens is not wholly accidental. The stars here seem to form a system, and a system which one can hardly conceive to be wholly unconnected with the neighbouring stream of the Milky Way. But in the southern portion the arrangement is yet more remarkable and significant. From Scorpio, over the feet of the Centaur, over the keel of Argo, to Canis Major, there is a clustering of brilliant stars, which it seems wholly impossible not to connect with the background of nebulous light. It is noteworthy, also, that this stream of stars merges into the stream commencing with the group of Orion already noticed. Nor is this all. It is impossible not to be struck by the marked absence of stars in that region of the sky which lies in the upper right-hand corner of Fig. 2. One has the impression that the stars have been attracted towards the region of the stream indicated, so as to leave this space comparatively bare.

Now, this last circumstance would appear less remarkable if the paucity of stars here noticed were common also in parts of the heavens far removed from the Milky Way. But this is not the case. Beyond this very region, which we find so bare of stars, we come upon a region in which stars are clustered in considerable density, a region including Crater, Corvus, and Virgo, with the conspicuous stars Algores, Alkes, and Spica. But, what is very remarkable, while we can trace a connection between the stream of bright stars in Fig. 2, and the stream

of nebulous light in the background, it is obvious that the two streams are not absolutely coincident in direction. The stream lies (in the figure) above the Milky Way near Scorpio, crosses it in the neighbourhood of Crux, and passes below it along Canis Minor, Orion, and Taurus. Does the stream return to the Milky Way? It seems to me that there is clear evidence of a separation near Aldebaran, one branch curving through Auriga, Perseus, and Cassiopeia, the other proceeding (more nearly in the direction originally observed) through Aries (throwing out an outlier along the band of Pisces), over the square of Pegasus, and along the streams which the ancients compared to water from the urn of Aquarius (but which in our modern maps are divided between Aquarius and Grus). The stream-formation here is very marked, as is evident from the phenomenon having attracted the notice of astronomers so long ago. But modern travels have brought within our ken the continuation of the stream over Toucan, Hydrus, and Reticulum (the two latter names being doubtless suggested by the convolutions of the stream in this neighbourhood). Here the stream seems to end in a sort of double loop, and it is not a little remarkable that the Nubecula Major lies within one loop, the Nubecula Minor within the other. It is also noteworthy that from the foot of Orion there is another remarkable stream of stars, recognised by the ancients under the name of the River Eridanus, which proceeds in a sinuous course towards this same region of the Nubecula.

Having thus met with evidence-striking at least, if not decisive, of a tendency to aggregation into streams, let us consider if, in other parts of the heavens, similar traces may not be observable. We traced a stream from Scorpio towards Orion, and so round in a spiral to the Nubeculæ. Let us now return to Scorpio, and trace the stream (if any appear) in the contrary direction. Now although over the northern hemisphere starstreams are not nearly so marked as over the southern, yet there appears a decided indication of stream-formation along Serpens and Corona over the group on the left hand of Bootes to the Great Bear. A branch of this stream, starting from Corona, traverses the body of Bootes, Berenice's Hair, the Sickle in Leo, the Beehive in Cancer, passing over Castor and Pollux in Gemini, towards Capella. A branch from the feet of Gemini passes over Canis Minor, along Hydra (so named doubtless from the obvious tendency to stream-formation along the length of this constellation), and so to the right claw of Scorpio. Four small stars of Hydrus are indicated in Fig. 2 between Scorpio and Centaurus near the upper edge of the figure.

One other remarkable congeries of stars is to be mentioned.

From the northern part of the Milky Way there will be noticed a projection towards the north pole from the head of Cepheus. This projection seems to merge itself in a complex convolution of stars forming the ancient constellation Draco, which doubtless included the ancient (but probably less ancient) constellation Ursa Minor. After following the convolutions of Draco, we reach the bright stars Alwaid and Etanin (8 and y) of this constellation, and thence the stream passes to Lyra, where it seems to divide into two, one passing through Hercules, the other along Aquila, and curving upwards (see Fig. 1) into the remarkable group Delphinus.

The streams here considered, include every conspicuous star in the heavens. But the question will at once suggest itself, whether we have not been following a merely fanciful scheme, whether all these apparent streams might not very well be supposed to result from mere accident. Now, from experiments I have made, I am inclined to believe that in any chance distribution of points over a surface, the chance against the occurrence of a single stream so marked as that which lies (in part) along the back of Grus, or (to take one within our figure) as the curved stream of bright stars along Scorpio, is very great indeed. I am certain that the occurrence of many such streams is altogether improbable. And wherever one observes a tendency to stream-formation in objects apparently distributed wholly by chance, one is led to suspect, and thence often to detect the operation of law. I will take an illustration, very homely perhaps, but which will serve admirably to explain my meaning. In soapy water, left in a basin after washing, there will often be noticed a tendency to the formation of spiral whorls on the surface. In other cases there may be no spirality, but still a tendency to stream-formation. Now, in this case, it is easy to see, that the curved bottom of the basin has assisted to generate streams in the water, either circulating in one direction, or opposing and modifying each other's effects, according to the accidental character of the disturbance given to the water in the process of washing. Here, of course, there can be no doubt of the cause of the observed phenomena; and I believe that in every case in which even a single marked stream is seen in any congeries of spots or points, a little consideration will suggest a regulating cause to which the peculiarity may be referred.

It is hardly necessary to say that, if the stream-formation

Sometimes, a singular regularity of curvature is noticed, and a spiral is formed closely resembling in configuration some of the great spiral nebulæ, as drawn by Lord Rosse, so that one is tempted to see in the centrifugal tendency of the disturbed water, and the centripetal effects caused by reflection from the basin's surface, causes which may in some sense illustrate the laws operating in wider domains of space.

I have indicated is considered to be really referable to a systematic distribution, the theory of a stratum of stars distributed with any approach to uniformity, either as respects magnitude or distance must be abandoned. It seems to me to be also quite clear that the immense extent of the galaxy as compared with the distances of the 'lucid' stars from us, could no longer be maintained. On this last point we have other evidence, which I will briefly consider.

First, there is the evidence afforded by clusterings in the Milky Way. I will select one which is well known to every telescopist, namely, the magnificent cluster on the sword-hand of Perseus. No doubt can be entertained that this cluster belongs to the galactic nebula, that is, that it is not an external cluster: the evidence from the configuration of the spot and from the position it occupies, is conclusive on this point. Now, within this spot, which shows no stars to the naked eye, a telescope of moderate power reveals a multitude of brilliant stars, the brightest of which are of about the seventh magnitude. Around these there still appears a milky unresolved light. If a telescope of higher power be applied, more stars are seen, and around these there still remains a nebulous light. Increase power until the whole field blazes with almost unbearable light, yet still there remains an unresolved background. "The illustrious Herschel" says Professor Nichol, "penetrated, on one occasion, into this spot, until he found himself among depths, whose light could not have reached him in much less than 4000 years; no marvel that he withdrew from the pursuit, conceiving that such abysses must be endless." It is precisely this view that I wish to controvert. And I think it is no difficult matter to show at least a probability against the supposition that the milky light in the spot is removed at a vast distance behind the stars of the seventh magnitude seen in the same field.

The supposition amounts, in fact, to the highly improbable view that we are looking here at a range of stars extending in a cylindrical stratum directly from the eye-a stratum whose section is so very minute in comparison with its breadth, that, whereas the whole field within which the spot is included is but small, the distance separating the nearest parts of the group from the farthest, is equivalent to the immense distance supposed to separate the sphere of seventh magnitude stars from the extreme limits of our galaxy. And the great improbability of this view is yet further increased, when it is observed that within this spot there is to be seen a very marked tendency to the formation of minor streams, around which the milky light seems to cling. It seems, therefore, wholly improbable that the cluster really has that indefinite longitudinal extension suggested by Professor Nichol; I think, therefore, that the