regions I., II., and III., and also in VII., VIII., and IX., but would suddenly increase in IV. and VI. as the boundary of the ring was approached. Instead of such being the case, the numbers 2.78, 3.03, and 3.54 in the north, and 3.14, 3.21, and 3.71 in the south, show a progressive increase from the galactic pole to the Galaxy itself. The conclusion to be drawn is a fundamental one. The universe, or at least the denser portion of it, is really flattened between the galactic poles, as supposed by Herschel and Struve.'
But looking at the series of figures in the table, and again as quoted by Professor Newcomb, they seem to me to show in some measure what he says they do not show. I therefore drew out the above diagram from the figures in the table, and it certainly shows that the density in regions I., II., and III., and in regions VII., VIII., and IX., may be said to be 'about the same,' that is, they increase very slowly, and that they do 'suddenly increase' in IV. and VI. as the boundary of the Galaxy is approached. This may be explained either by a flattening towards the poles of the Galaxy, or by the thinning out of stars in that direction.
In order to show the enormous difference of star-density in the Galaxy and at the galactic poles, Professor Newcomb gives the following table of the Herschelian gauges, on which he only remarks that they show an enormously increased density in the galactic region due to the Herschels having counted so many more stars there than any other observers.
DIAGRAM OF STAR-DENSITY
From a table in The Stars (p. 249).
But an important characteristic of these figures is, that the Herschels alone surveyed the whole of the heavens from the north to the south pole, that they did this with instruments of the same size and quality, and that from almost life-long experience in this particular work they were unrivalled in their power of counting rapidly and accurately the stars that passed over each field of view of their telescopes. Their results, therefore, must be held to have a comparative value far above those of any other observer or combination of observers. I have therefore thought it advisable to draw a diagram from their figures, and it will be seen how strikingly it agrees with the former diagram in the very slow increase of star-richness in the first three regions north and south, the sudden increase in regions IV. and VI. as we approach the Galaxy, while the only marked difference is in the enormously greater richness of the Galaxy itself, which is an undoubtedly real phenomenon, and is brought out here by the unrivalled observing power of the two greatest astronomers in this special department that have ever lived.
We shall find later on that Professor Newcomb himself, as the result of a quite different inquiry arrives at a result in accordance with these diagrams which will then be again referred to. As this is a very interesting subject, it will be well to give another diagram from two tables of star-density in Sir John Herschel's volume already quoted. The tables are as follows:—
In these tables the Milky Way itself is taken as occupying two zones of 15° each, instead of one of 20° as in Professor Newcomb's tables, so that the excess in the number of stars over the other zones is not so large. They show also a slight preponderance in all the zones of the southern hemisphere, but this is not great, and may probably be due to the clearer atmosphere of the Cape of Good Hope as compared with that of England.
DIAGRAM OF STAR-DENSITY.
From Table in Sir J. Herschel's Outlines of Astronomy (10th ed., pp. 577-578).
It need only be noted here that this diagram shows the same general features as those already given, of a continuous increase of star-density from the poles of the Galaxy, but more rapidly as the Galaxy itself is more nearly approached. This fact must, therefore, be accepted as indisputable.
An important factor in the structure of the heavens is afforded by the distribution of the two classes of objects known as clusters and nebulæ. Although we can form an almost continuous series from double stars which revolve round their common centre of gravity, through triple and quadruple stars, to groups and aggregations of indefinite extent—of which the Pleiades form a good example, since the six stars visible to the naked eye are increased to hundreds by high telescopic powers, while photographs with three hours' exposure show more than 2000 stars—yet none of these correspond to the large class known as clusters, whether globular or irregular, which are very numerous, about 600 having been recorded by Sir John Herschel more than fifty years ago. Many of these are among the most beautiful and striking objects in the heavens even with a very small telescope or good opera-glass. Such is the luminous spot called Praesepe, or the Beehive in the constellation Cancer, and another in the sword handle of Perseus.
In the southern hemisphere there is a hazy star of about the fourth magnitude, Omega Centauri, which with a good telescope is seen to be really a magnificent cluster nearly two-thirds the diameter of the moon, and described by Sir John Herschel as very gradually increasing in brightness to the centre, and composed of innumerable stars of the thirteenth and fifteenth magnitudes, forming the richest and largest object of the kind in the heavens. He describes it as having rings like lace-work formed of the larger stars. By actual count, on a good photograph, there are more than 6000 stars, while other observers consider that there are at least 10,000. In the northern hemisphere one of the finest is that in the constellation Hercules, known as 13 Messier. It is just visible to the naked eye or with an opera glass as a hazy star of the sixth magnitude, but a good telescope shows it to be a globular cluster, and the great Lick telescope resolves even the densest central portion into distinct stars, of which Sir John Herschel considered there were many thousands. These two fine clusters are figured in many of the modern popular works on astronomy, and they afford an excellent idea of these beautiful and remarkable objects, which, when more thoroughly studied, will probably aid in elucidating some of the obscure problems connected with the constitution and development of the stellar universe.
But for the purpose of the present work the most interesting fact connected with star-clusters is their remarkable distribution in the heavens. Their special abundance in and near the Milky Way had often been noted, but the full importance of the fact could not be appreciated till Mr. Proctor and, later, Mr. Sidney Waters marked down, on maps of the two hemispheres, all the star-clusters and nebulæ in the best catalogues. The result is most interesting. The clusters are seen to be thickly strewn over the entire course of the Milky Way, and along its margins, while in every other part of the heavens they are thinly scattered at very distant intervals, with the one exception of the Magellanic clouds of the southern hemisphere where they are again densely grouped; and if anything were needed to prove the physical connection of these clusters with the Galaxy it would be their occurrence in these extensive nebulous patches which seem like outlying portions of the Milky Way itself. With these two exceptions probably not one-twentieth part of the whole number of star-clusters are found in any part of the heavens remote from the Milky Way.
Nebulæ were for a long time confounded with star-clusters, because it was thought that with sufficient telescopic power they could all be resolvable into stars as in the case of the Milky Way itself. But when the spectroscope showed that many of the nebulæ consisted wholly or mainly of glowing gases, while neither the highest powers of the best telescopes nor the still greater powers of the photographic plate gave any indications of resolvability, although a few stars were often found to be, as it were, entangled in them, and evidently forming part of them, it was seen that they constituted a distinct stellar phenomenon, a view which was enforced and rendered certain by their quite unique mode of distribution. A few of the larger and irregular type, as in the case of the grand Orion nebula visible to the naked eye, the great spiral nebula in Andromeda, and the wonderful Keyhole nebula round Eta Argûs, are situated in or near the Milky Way; but with these and a few other exceptions the overwhelming majority of the smaller irresolvable nebulæ appear to avoid it, there being a space almost wholly free from nebulæ along its borders, both in the northern and southern hemispheres; while