Astronomy

The spiral nebulæ are wonderful objects, and were discovered by the late Lord Rosse, with his great six-foot telescope. Their character has been fully confirmed by photographs taken by Dr. Roberts. One of the most remarkable of these extraordinary objects is that known as 51 Messier. It lies about three degrees south-west of the bright star Eta Ursæ Majoris—the star at the end of the Great Bear’s tail. It was discovered by Messier while comet-hunting on October 13, 1773. Telescopes of moderate power merely show two nebulæ nearly in contact, but Lord Rosse saw it as a wonderful spiral, and his drawing agrees fairly well with a photograph taken by Dr. Roberts in April, 1889. The nebula has also been photographed by Dr. Common. Dr. Roberts says: “The photograph shows both nuclei of the nebula to be stellar, surrounded by dense nebulosity, and the convolutions of the spiral in this as in other spiral nebulæ are broken up into star-like condensations with nebulosity around them. Those stars that do not conform to the trends of the spiral have nebulous trails attached to them, and seem as if they had broken away from the spirals.” A tendency to a spiral structure in the smaller nebula is also visible on the original negative. Dr. Huggins finds that the spectrum is not gaseous.

The nebulæ known as 99 Messier is of the spiral form. It lies on the borders of Virgo and Coma Berenices, near the star 6 Comæ. In large telescopes it somewhat resembles a “Catherine wheel.” D’Arrest and Key thought it resolvable into stars. It has been photographed by M. Von Gothard.

Among the clusters and nebulæ, we may class the Magellanic Clouds, or Nubeculæ in the Southern Hemisphere, as they consist of stars, clusters, and nebulæ. These very remarkable objects form two bright spots of milky light, which, at first sight, look like luminous patches of the Milky Way, but are in no way connected with the Galaxy. Sir John Herschel, speaking of the larger cloud, says: “The immediate neighbourhood of the Nubecula Major is somewhat less barren of stars than that of the Minor, but it is by no means rich, nor does any branch of the Milky Way whatever form any certain or conspicuous junction with, or include, it,” and again he says, with reference to the smaller cloud: “Neither with the naked eye, nor with a telescope, is any connexion to be traced either with the greater Nubecula, or with the Milky Way.” The Nubeculæ are roughly circular in form, and, viewed with the naked eye, they very much resemble irresolvable nebulæ as seen in a telescope. The larger cloud, or Nubecula Major, as it is called, is of considerable extent, and covers about 42 square degrees, or over two hundred times the apparent size of the full moon. It was called by the Arabs el-baker, or “the White Ox,” and is referred to by Al-Sûfi in his “Description of the Heavens,” written in the tenth century. When examined with a good telescope, it is found to consist of about six hundred stars of the sixth to the tenth magnitude, with many fainter ones, and about three hundred clusters and nebulæ. Sir John Herschel, in his “Cape Observations,” says: “The Nubeculæ Major, like the Minor, consists partly of large tracts and ill-defined patches of irresolvable nebula, and of nebulosity in every stage of resolution, up to perfectly resolved stars like the Milky Way, as also of regular and irregular nebulæ properly so-called, of globular clusters in every stage of resolvability, and of clustering groups sufficiently insulated and condensed to come under the designation of ‘clusters of stars.’... It is evident, from the intermixture of stars and unresolved nebulosity, which probably might be resolved with a higher optical power, that the nubeculæ are to be regarded as systems sui generis, and which have no analogues in our hemisphere.”

The smaller Magellanic Cloud, or Nubecula Minor, is fainter to the eye, and not so rich in the telescope. It covers about 10 square degrees, or about fifty times the area of the full moon. Sir John Herschel, in his “Cape Observations,” describes it as “a fine large cluster of very small stars, 12 ... 18 magnitude, which fills more than many fields, and is broken into many knots, groups, and straggling branches, but the whole (i.e., the whole of the clustering part) is clearly resolved.” It is surrounded by a barren region remarkably devoid of stars. Sir John Herschel says: “The access to the Nubecula Minor is on all sides through a desert.”... “It is preceded at a few minutes in R. A. by the magnificent globular cluster, 47 Toucani (Bode), but is completely cut off from all connexion with it; and with this exception, its situation is in one of the most barren regions in the heavens.” Herschel found the middle of the cloud clearly resolved into stars, while its edges remained irresolvable with his large reflector. He says: “The edge of the smaller cloud comes on as a mere nebula.... We are now in the cloud. The field begins to be full of a faint light perfectly irresolvable.... I should consider about this place to be the body of the cloud which is here fairly resolved into excessively minute stars.... It is not like the stippled ground of the sky. The borders fade away, quite insensibly, and are less or not at all resolved.” Herschel gives a catalogue of 244 objects in the Nubecula Minor. Of these about 200 are stars, and the remainder nebula and clusters. From this it appears that the smaller nubecula contains a much larger proportion of stars than the larger cloud.

Judging from their roughly globular form, the dimensions of the Magellanic Clouds are probably small compared with their distance from the earth, so that in these remarkable objects—particularly in the larger cloud—we see stars of the seventh, eighth, ninth, and tenth magnitude, apparently mixed up with fainter stars, and “clusters of all degrees of resolvability,” and Sir John Herschel says: “It must therefore be taken as a demonstrated fact, that stars of the seventh or eighth magnitude, and irresolvable nebulæ, may co-exist within limits of distance not differing in proportion more than as 9 to 10.”^[140] It should be remembered, however, that possibly some of the fainter stars may—as in the Pleiades—lie far out in space beyond the greater Magellanic Cloud.

The Magellanic Clouds have recently been photographed by Mr. Russell at the Sydney Observatory. He finds the larger cloud—the Nubecula Major—to be of a most complex form, with evidence of a spiral structure, a feature also traceable, but not so clearly, in a photograph of the Nubecula Minor, or smaller cloud.

Dr. Dreyer’s new index catalogue of recent discoveries of nebulæ, together with the general catalogue previously published, gives the position of 9,369 nebulæ.^[141] A very small proportion of the new discoveries have been made by photography, and more than half of them were found by M. Javelle with the great refractor of the Nice Observatory. Most of the new objects are very small and faint, and form probably “only a small portion of the number visible in large telescopes.”

Several nebulæ have been suspected of variation in light. One discovered by Dr. Hind in 1852 near the variable star T Tauri was found to be an easy object with the great Lick telescope in February, 1895, but in September of the same year it had “entirely vanished.” In the same instrument, “T Tauri was involved in a small hazy nebulosity, but the definite nebula in which it shone in 1890 did not exist in September, 1895.”^[142]

CHAPTER VII.
THE CONSTRUCTION OF THE HEAVENS.

The construction of the visible universe is one of great interest, but of considerable difficulty. If we reflect that in viewing the starry heavens we are placed at the centre of a hollow sphere of indefinite extent, and that the distance of only a few of the stars from the earth has hitherto been ascertained with any approach to accuracy, the great difficulty of framing a satisfactory theory of the construction of the heavens will be easily understood.

In considering the subject, let us first inquire as to the probable number of stars visible in our largest telescopes. Are the visible stars infinite or limited in number? The reply to this question is easy. As the number of stars visible to the naked eye is limited, so the number of stars visible in the largest telescopes is limited also. Those who do not give the subject sufficient consideration seem to think that the number of the stars is practically infinite, or at least that the number is so great that it cannot be estimated. But this idea is totally incorrect, and due to complete ignorance of telescopic revelations. It is certainly true that, to a certain extent, the larger the telescope used in the examination of the heavens, the more the number of the stars seems to increase; but we now know that there is a limit to this increase of telescopic vision. And the evidence clearly shows that we are rapidly approaching this limit. Although the number of stars visible in the Pleiades rapidly increases at first with increase in the size of the telescope used, and although photography has still further increased the number of stars in this remarkable cluster, it has recently been found that an increased length of exposure—beyond three hours—adds very few stars to the number visible on the photograph taken at the Paris Observatory in 1885, on which over 2,000 stars can be counted. Even with this great number on so small an area of the heavens, comparatively large vacant spaces are visible between the stars, and a glance at the original photograph is sufficient to show that there would be ample room for many times the number actually visible. I find that, if the whole heavens were as rich in stars as the Pleiades, there would be only 33 millions in both hemispheres.

On a photograph of the region surrounding Gamma Cassiopeiæ, taken by Dr. Roberts in December, 1895, with a reflecting telescope of 20 inches aperture, and an exposure of two hours and twelve minutes, he finds 17,100 stars on an area of four square degrees. This would give for the whole area of the heavens—if equally rich in stars—a total of about 176 millions; but Gamma Cassiopeiæ lies in a rich region of the Milky Way, and probably the great majority of the stars shown on Dr. Roberts’ photograph belong to the Galaxy, which we know to be especially rich in stars. One thing is certain, that the heavens as a whole are not nearly so rich as this particular spot. There may, perhaps, be richer spots elsewhere in the Milky Way, but in other parts of the sky there are many regions considerably poorer.

Let us consider a still more extreme case of stellar richness. On a photograph of the great globular cluster, Omega Centauri, recently taken in Peru, a count of the stars has been carefully made by Professor and Mrs. Bailey, and, as stated in the last chapter, the number of stars contained in the cluster may be taken as 10,000. Now, if the whole sky were as thickly studded with stars as in this cluster, the total number visible in the whole heavens would be 1,650 millions, a very large number, of course, but not much in excess of the present population of the earth, and I am not aware that the number of the earth’s inhabitants has ever been described as “infinite.”

Clusters, such as the Pleiades and Omega Centauri, are, of course, remarkable, and rare exceptions to the general rule of stellar distribution, and the heavens in general are not—even in the richest portions of the Milky Way—nearly so rich in stars as the globular clusters. The fact of these clusters being remarkable objects, proves that they are unusually rich in stars, and there is strong evidence—evidence amounting to absolute proof in the case of the globular clusters—that these collections of stars are really, and not apparently, close, and that they are actually systems of suns, and occupy a comparatively limited volume in space. We cannot, then, estimate the probable number of the visible stars by counting those visible in one of the globular clusters.

That the number of the visible stars will not probably be largely increased by any increase in telescopic power, is indicated by the fact that Celoria, using a small telescope, of power barely sufficient to show stars to the eleventh magnitude, found that he could see almost exactly the same number of stars near the north pole of the Milky Way as were visible in Sir William Herschel’s great telescope! thus indicating that, here at least, no increase of optical power will materially increase the number of stars visible in that direction; for Herschel’s large telescope certainly showed far fainter stars than those of the eleventh magnitude in other portions of the heavens. It should therefore have shown fainter stars at the pole of the Milky Way also, if such stars existed in that region of space. Their absence, therefore, seems certain proof that very faint stars do not exist in that direction, and that, here at least, our sidereal universe is limited in extent A photograph, taken by Dr. Roberts not very far from the spot in question, shows only 178 stars to the square degree. This rate of distribution would give a total of only 7,343,000 stars for both hemispheres!

An examination by Miss Clerke of Professor Pickering’s catalogue of stars surrounding the north pole of the heavens shows that “the small stars are overwhelmingly too few for the space they must occupy, if of average brightness; and they are too few in a constantly increasing ratio.”^[143] Here again, a “thinning out” of the stellar hosts seems clearly indicated, and suggests that a limit will soon be reached, beyond which our most powerful telescopes and photographic plates will fail to reveal any further stars.

Let us now consider the number of stars actually visible. Maps of the northern portion of the heavens have been published by Argelander and Heis, and charts of the southern sky by Behrmann and Gould. Heis shows stars to about magnitude 6⅓, and Behrmann to about the same brightness. I find that the total number shown by both observers, as visible to the naked eye, is 7,249. The total number, to the sixth magnitude inclusive, shown by both observers, is 4,181. Argelander gives 5,000 stars to the sixth magnitude inclusive, and for stars to the ninth magnitude, the following numbers in each magnitude:—First magnitude, 20; second magnitude, 65; third magnitude, 190; fourth magnitude, 425; fifth magnitude, 1,100; sixth magnitude, 3,200; seventh magnitude, 13,000; eighth magnitude, 40,000; and ninth magnitude, 142,000, or a total of “200,000 for the entire number of stars from the first to the ninth magnitude inclusive.”^[144] This result agrees closely with an estimate previously made by Struve. From a formula given by Dr. Gould, deduced from observations in the Southern Hemisphere, I find the number of stars to the ninth magnitude inclusive would be 215,674, so that Argelanders estimate of 200,000 stars to the ninth magnitude inclusive cannot be far from the truth. It will be seen from Argelanders figures that the number of stars in each class of magnitude is roughly three times that in the class one magnitude brighter. Supposing this progressive increase continued to the seventeenth magnitude—the faintest visible in the great Lick telescope—I find that the total number of stars would be nearly 1,400 millions, or less than the number found from a consideration of the cluster Omega Centauri. But it is evident from Celoria’s observation, referred to above, and from Professor Pickering’s photographs of stars near the North Pole, that the fainter stars do not increase in the ratio assumed above. We must therefore conclude that there is a “thinning out” of the fainter stars at some point below the ninth magnitude. Taking into consideration the rich regions of the Milky Way, and the comparatively poor portions of the sky, it is now generally admitted by astronomers, who have studied this particular question, that the probable number of stars visible in our largest telescopes does not exceed 100 millions, a number which, large as it absolutely is, may be considered as relatively very small, and even utterly insignificant, when compared with an “infinite number.”

Let us see what richness of stellar distribution is implied by this number of 100 millions of visible stars. It may be easily shown that the area of the whole sky, in both hemispheres, is 41,253 square degrees, or about 200,000 times the area of the full moon. This gives 2,424 stars to the square degree. The moon’s apparent diameter being slightly over half a degree (31′ 5″), the area of its disc is about one-fifth of a square degree. Hence, for 100 millions of stars in the whole star sphere, we have 485 stars to each space of sky, equal in area to the full moon. This seems a large number, but stars scattered even as thickly as this would appear at a considerable distance apart when viewed with a large telescope and a high power. As the area of the moon’s disc contains about 760 square minutes of arc, there would not be an average of even one star to each square minute. A pair of stars half a minute, or 30 seconds, apart, would form a very wide double star, and with stars placed at even this distance, the moon’s disc would cover about 3,000, or over six times the actual number visible in the largest telescopes. In Dr. Roberts’ photograph of the region surrounding Gamma Cassiopeiæ, which shows over 17,000 stars, on four square degrees, or over 4,000 stars to the square degree, the stars do not seem very crowded, and there is a good deal of black sky visible between them.

But, in addition to the conclusive evidence as to the limited number of the visible stars derived from actual observation and the results of photography, we have indisputable evidence from mathematical considerations that the number of the visible stars must necessarily be limited. For were the stars infinite in number, and scattered through infinite space with any approach to uniformity, it may be proved that the whole heavens would shine with the brightness of the sun. As the surface of a sphere varies as the square of its radius, and light inversely as the square of the distance (or radius of the star sphere at any point), we have the diminished light of the stars exactly counterbalanced by the increased number at any given distance. For a distance of say ten times the distance of the nearest fixed star, the light of each star would be diminished by the square of 10 or 100 times, but the total number of stars would be 100 times greater, so that the total star light would be the same. This would be true for all distances. The total light would therefore—by addition—be proportional to the distance, and hence, for an infinite distance we should have an infinite amount of light For an infinite number of stars, therefore, we should have a continuous blaze of light over the whole surface of the visible heavens. Far from this being the case, the amount of light afforded by the stars on the clearest nights is, on the contrary, comparatively small, and the blackness of the background, “the darkness behind the stars,” is very obvious. According to Miss Clerke (“System of the Stars,” p. 7), the total light of all the stars, to magnitude 9½, is about one-eightieth of full moonlight. M. G. l’Hermite found for the total amount of starlight one-tenth of moonlight; but this estimate is evidently too high. Assuming the sun’s brightness as 28 magnitudes brighter than a star of the first magnitude,^[145] and Zöllner’s estimate that sunlight is 618,000 times that of moonlight, I find that the total light of the stars to magnitude 9½, as stated by Miss Clerke, would be equivalent to the combined light of about 320,000 stars of the sixth magnitude, or 3,200 stars of the first magnitude. Even taking M. l’Hermite’s high estimate of one-tenth of moonlight, the total starlight would be represented by 25,600 stars of the first magnitude.

To explain the limited number of the visible stars, several hypothesis have been advanced. If space be really infinite, as we seem compelled to suppose, it would be reasonable to expect that the number of the stars would be practically infinite also. But, as I have shown above, the number of the visible stars is certainly finite, and the number visible and invisible must be finite also, for otherwise the amount of starlight would be much greater than it is. To account for the limited number of visible stars, it has been suggested that beyond a certain distance in space, there may be an “extinction of light,” caused by absorption in the luminiferous ether. In a recent paper on this subject, Schiaparelli, the famous Italian astronomer, suggests that if any extinction of light really takes place, it may probably be due, not to absorption in the ether, but to fine particles of matter scattered through interstellar space. In support of this hypothesis, he refers to the supposed constitution of comets’ tails, of falling stars, and meteorites, and he shows that the quantity of matter necessary to produce the required extinction would be very small—so small, indeed, that a quantity of this matter scattered through a volume equal to that of the earth, if collected into one mass, would only form a ball of less than one inch in diameter. We can readily admit the existence of such a minute quantity of matter in a fine state of subdivision scattered through space, but it seems to me much more probable that the limited number of the visible stars is due, not to any extinction of their light by absorption in the ether, or by fine particles scattered through space, but to a real thinning out of the stars as we approach the limits of our sidereal universe. Celoria’s observation, mentioned above, seems to prove that near the pole of the Milky Way very few stars fainter than the eleventh magnitude are visible, even in a large telescope, and Dr. Roberts’ photographs, taken in the vicinity of the celestial pole, confirm this conclusion. Now, this paucity of stars of the fainter magnitudes cannot be due to any absorption of light in the ether, for numerous stars of the sixteenth magnitude, or perhaps fainter, are visible in other parts of the heavens, and if in one place, why not in another? Sir John Herschel’s observations of the Milky Way in the Southern Hemisphere appear to render the hypothesis of any extinction of light very improbable. He says that the hypothesis, “if applicable to any, is equally so to every part of the Galaxy. We are not at liberty to argue that at one part of its circumference our view is limited by this sort of cosmical veil, which extinguishes the smaller magnitudes, cuts off the nebulous light of distant masses, and closes our view in impenetrable darkness; while at another we are compelled, by the clearest evidence telescopes can afford, to believe that star-strewn vistas lie open, exhausting their powers, and stretching out beyond their utmost reach, as is proved by that very phænomenon which the existence of such a veil would render impossible, viz., infinite increase of number and diminution of magnitude, terminating in complete irresolvable nebulosity.”

How then are we to explain the limited number of the visible stars? If space be infinite, as we seem compelled to suppose, the number of the stars would probably be infinite also, or at least vastly greater than the number actually visible. It has been suggested that, owing to the progressive motion of light, the light of very distant stars may probably not yet have reached the earth, although travelling through space for thousands of years. But considering the vast periods of time during which the stellar universe has probably been in existence, this hypothesis seems very unsatisfactory. The most probable hypothesis seems to be that all the stars, clusters and nebulæ, visible in our largest telescopes, form together one vast system, which constitutes our visible universe, and that this system is isolated by a starless void from other similar systems which probably exist in infinite space. The distance between these separate systems—or “island universes,” as they have been called—may be very great, compared with the diameter of each system, in the same way that the diameter of our visible universe is very great compared with the diameter of the solar system. As the sun is a star, and the stars are suns, and as our sun is separated from his neighbour suns in space by a sunless void, so may our universe be separated from other universes by a vast and starless abyss. On this hypothesis, the supposed extinction of light—which may have little or no perceptible effect within the limits of our visible universe—may possibly come into play across the vast and immeasurable distances which probably separate the different universes from each other, and may perhaps extinguish their light altogether.

Another hypothesis which also seems possible is that the luminiferous ether which extends throughout our visible universe may perhaps be confined to this universe itself, and that beyond its confines, the ether may thin out, as our atmosphere does at a certain distance from the earth, and finally cease to exist altogether, ending in an absolute vacuum, which would, of course, arrest the passage of all light from outer space, and thus produce “the darkness behind the stars.”

Let us now consider the apparent distribution of the stars and nebulæ on the celestial vault, and their probable relation to each other in space. As already stated, Argelander considered the number of stars of the first magnitude to be about twenty, but modern photometric measures have reduced this number to thirteen or fourteen. According to the Harvard measures, the fourteen brightest stars in the heavens, in order of magnitude, are: Sirius, Canopus, Arcturus, Capella, Vega, Alpha Centauri, Rigel, Procyon, Achernar, Beta Centauri, Betelgeuse, Altair, Aldebaran and Alpha Crucis. Seven of these are in the Northern Hemisphere, namely: Arcturus, Capella, Vega, Procyon, Betelgeuse, Altair, and Aldebaran; and seven in the Southern Hemisphere: Sirius, Canopus, Alpha Centauri, Rigel, Achernar, Beta Centauri, and Alpha Crucis, so that the brightest stars are pretty evenly distributed between the two hemispheres. Of these bright stars, no less than twelve lie in or near the Milky Way, Arcturus and Achernar being the only two at any considerable distance from the Galaxy. This is very remarkable and suggestive, as the area covered by the Milky Way is probably not more than one-fourth of the whole star sphere.

Of the stars fainter than the first magnitude, but brighter than magnitude 2·0, there are about 10 in the Northern Hemisphere, of which 4 lie in or near the Milky Way, and about 19 in the Southern Hemisphere, of which no less than 14 are situated in or near the Galaxy.

Of those brighter than magnitude 3·0, I find 33 stars in or near the Milky Way out of a total of about 95 in both hemispheres. To extend this investigation to all stars visible to the naked eye, I made, some years since, an examination of all the stars in Heis’ atlas that lie in the Milky Way, and found that number to be 1,186 out of a total of 5,356, or a percentage of about 22. At my request, Col. Markwick, F.R.A.S., made a similar count for the stars in Dr. Gould’s charts of the Southern Hemisphere (Uranometria Argentina), and found that, down to the fourth magnitude, there are 121 stars on the Milky Way out of 228, or a percentage of 53, and for all stars to the seventh magnitude inclusive, there are 3,072 on the Milky Way out of a total of 6,694, or a percentage of nearly 46. Col. Markwick finds that the Milky Way in the Southern Hemisphere, as shown on Gould’s charts, covers about one-third of the whole hemisphere. As will be seen by the above figures, the percentage of stars, even to the fourth magnitude, lying on the Milky Way is considerably greater than this proportion.

The above results show that the brighter stars which are apparently projected on the Milky Way probably belong to that zone, and are not merely fortuitously scattered over the surface of the heavens.

To extend the investigation still further, and include stars to the eighth magnitude, I made an examination of the stars shown on Harding’s charts to that magnitude, in a zone of 30° in width—15° degrees on each side of the Equator—and found a marked increase in the number of stars where the zone crossed the Milky Way. The numbers per hour of Right Ascension varied from a minimum of 275 (hours I. and II.) to maxima of 601 in the Milky Way in Monoceros, and 611 in the Galaxy in Serpens and Aquila. A valuable investigation by the late Mr. Proctor went further still. He plotted all the stars shown in the charts of Argelander’s Durchmusterung, which contains stars to 9½ or 10th magnitude. In this remarkable chart the course of the Milky Way is clearly defined by a marked increase of stellar density. Proctor says: “In the very regions where the Herschelian gauges showed the minutest telescopic stars to be most crowded, my chart of 324,198 stars shows the stars of the higher orders (down to the eleventh magnitude) to be so crowded that, by their mere aggregation within the mass, they show the Milky Way with all its streams and clusterings. This evidence, I venture to affirm, is altogether decisive as to the main question, whether large and small stars are really intermixed in many regions of space, or whether the small stars are excessively remote. It is utterly impossible that excessively remote stars could seem to be clustered exactly where relatively near stars are richly spread. This might happen, no doubt, in a single instance; but that it could be repeated over and over again, so as to account for all the complicated features seen in my chart of 324,198 stars, I maintain to be utterly incredible.”^[146]

From a careful examination of the Milky Way in Aquila and Cygnus, Mr. Easton finds that “(1) In the zones considered, the distribution of stars down to 9·5 magnitude corresponds to the greater or less intensity of galactic light. (2) There is a real correspondence of the general outlines of the galactic forms with the distribution of 11 magnitude stars, and with those of stars between 10 and 15 magnitude. (3) Thus, in general, for the zones considered, the faint stars which form the Milky Way are thickly or sparsely scattered in respectively the same regions as the stars in Argelander’s last class; it follows, therefore, with a great degree of probability, that there is a real connexion between the distribution of 9 and 10 magnitude stars and that of the very faint stars of the Milky Way. Consequently, the very faint stars are at a distance which does not greatly exceed that of 9–10 magnitude stars. If stars of 13–15 magnitude were at their theoretical distance, there would be no reason why they should have the same apparent distribution in galactic latitude and longitude as 9–10 magnitude stars separated from them by enormous intervals.”^[147]

There are some regions in both hemispheres especially rich in naked eye stars. Of these the following may be mentioned in the Northern Hemisphere:—the region including the Pleiades, and Hyades in Taurus, the Northern portion of Orion, and the adjoining part of Gemini, the constellation Lyra, the northern portion of Cygnus, Cassiopeia’s Chair, and Coma Berenices. In the Southern Hemisphere there are several rich spots. A rich region extends from Canis Major to the Southern Cross, and nearly coincides with the course of the Milky Way. The richest spot of all, and perhaps the richest in the whole heavens in naked eye stars—with exception of the Pleiades—is that including the Southern Cross. This spot has an average of three stars to five square degrees, and if the whole heavens were as richly studded with stars there would be about 24,000 visible to the naked eye! The poverty of the adjoining “coal sack” is very remarkable. Another rich spot surrounds the variable star Eta Argûs, and the great nebula in Argo. There is another rich spot in the constellation Hydrus, not far from the greater Magellanic Cloud, and another will be found in Centaurus and Lupus, with its centre about Alpha of the latter constellation. According to Gould’s maps of the Southern Hemisphere, the richest region in stars down to the seventh magnitude is the southern portion of that part of the constellation Argo, known as Puppis.

In contrast to these rich regions, and in many cases closely adjoining them, are some barren regions, very poor in naked eye stars. For example, closely following the rich spot in Cassiopeia and between Iota Cassiopeiæ and Eta Persei is a remarkably poor spot, where a space of some sixty square degrees does not contain a single star brighter than the sixth magnitude! There is another poor region south of Alpha Hydræ, and another in the southern portion of the constellation Cetus.

A region of considerable extent, remarkably deficient in bright stars, will be noticed in the Northern Hemisphere. This comparatively barren region, which contains no star brighter than the fourth magnitude, is bounded by Cepheus, Cassiopeia, Perseus, Auriga, Gemini, Ursa Major, Draco, and Ursa Minor, and forms a conspicuous feature in the north-eastern portion of the sky in the early winter evenings. It will be noticed that the surrounding constellations all contain bright stars.

Whether the apparent crowding of stars in certain regions of the heavens is caused by a real proximity in space, or whether it is merely due to their being placed accidentally in the line of sight, is a question difficult to determine. In the case of star clusters, and especially the globular clusters, there is a high mathematical probability, amounting almost to absolute certainty, that they are comparatively close together, but in groups scattered over a considerable area, like those referred to above, the probability in favour of proximity is not so great. As we know the distance of so few stars from the earth, it is impossible to say whether the crowding is real or only apparent, but the probability seems to be that it is to some extent real.

A tendency to an arrangement of stars in streams was pointed out by Proctor in his “Universe and the Coming Transits.” This tendency to stream formation may be noticed on a large scale among the naked eye stars, for example, in Pisces, Scorpio, the River Eridanus, Aquarius, and the festoon of stars in Perseus. In some of these cases, of course, the stars are so far apart that the formation may be more apparent than real, but the tendency can also be clearly recognised among the fainter stars, and even among those only visible in telescopes and stellar photographs. This tendency to run in streams is well marked on the photographs taken at the Paris Observatory, and on those taken by Professor Barnard, Dr. Max Wolf, and others. It is a suggestive fact that these star streams are also very noticeable in star clusters, where there can be little or no doubt of a physical connexion between the component stars. With reference to a photograph of the southern portion of Aquila taken by Dr. Max Wolf in July, 1892, the late Mr. Ranyard, remarked: “Some of the streams of fainter stars in this region are very striking, and must convince the most sceptical of their reality. It is possible to draw an arc of a circle through any three stars, and a conic section through any five; but where we find ten or twenty stars falling into line, not once, but in many cases, and that there is a curious similarity between the strange curves and branching streams which these phalanges of stars mark out on the heavens, there is no room left for doubt that the mind is not being led away by a tendency of the imagination similar to that which finds faces in the fire, or sees a man carrying sticks on the face of the moon. If it is proved that a group of stars is arranged in line or marshalled in any order, it would follow that the individuals of the group must be actually as well as apparently close to one another, and that they form some kind of system, having all of them had a common origin, or been subject to some common influence.”^[148]

The great majority of the star clusters are found along the course of the Milky Way, while the irresolvable nebulæ seem to congregate towards the poles of the galactic zone.

Dr. Gould is of opinion that “a belt or stream of bright stars appears to girdle the heavens very nearly in a great circle, which intersects the Milky Way at about the points of its highest declination, and forms with it an angle not far from 20°; the southern node being near the margin of the Cross, and the northern in Cassiopeia.” According to Gould, this belt covers Orion, Canis Major, Columba, Puppis, Carina, the Southern Cross, Centaurus, Lupus, and the head of Scorpion in the Southern Hemisphere, its northern course being indicated by the brightest stars in Taurus, Perseus, Cassiopeia, Cepheus, Cygnus, and Lyra. Dr. Gould considers that our sun may possibly be a member of this belt of stars, which perhaps numbers less than 500, and which constitute “a small cluster, distinct from the vast organisation of that which forms the Milky Way, and of a flattened and somewhat bifid form. The southern portion of this supposed stream of bright stars had been previously recognised by Sir John Herschel, who says in his ‘Cape Observations,’ (p. 385), ‘It is about this region, or, perhaps, somewhat earlier, in the interval between η Argus and α Crucis, that the galactic circle, or medial line of the Milky Way may be considered as crossed by that zone of large stars, which is marked out by the brilliant constellation of Orion, the bright stars of Canis Major, and almost all the more conspicuous stars of Argo, the Cross, the Centaur, Lupus, and Scorpion. A great circle passing through ε Orionis and α Crucis will mark out the axis of the zone in question, whose inclination to the galactic circle is, therefore, about 20°, and whose appearance would lead us to suspect that our nearest neighbours in the sidereal system (if really such) form part of a subordinate sheet or stratum deviating to that extent from parallelism to the general mass which, seen projected on the heavens, forms the Milky Way.’”

These conclusions might seem probable enough when we compare the supposed zone of bright stars with the very diagrammatic drawings of the Milky Way as shown in many star maps; but when we consider the stars referred to with reference to the more artistic and accurate delineations of the Milky Way as drawn by Boeddicker, and even by Gould himself, we see that most of them are involved in the milky light of the Galaxy, and their connexion with the Milky Way itself seems quite as probable as that they form a belt distinct from the galactic zone. The apparent connexion of the stars in question with the Milky Way does not, however, disprove the existence of Dr. Gould’s belt or zone of bright stars. If the plane of the supposed belt nearly coincided with that of the Milky Way, the apparent connexion might not be real.

Mr. J. R. Sutton advances the theory^[149] that the Milky Way consists of “a great ring of large stars”—Dr. Gould’s solar cluster above referred to—“intersecting an equal ring of small ones (the Milky Way) at the extremities of a common diameter.” He considers that “the great star belt is a genuine girdle of stars in space, in which also the foundations of the sidereal system are laid, the Milky Way being an appendant to it of lesser rank.”

That the Milky Way really forms a ring of stars in space there is strong evidence to show. Sir William Herschel’s original theory that the galactic gleam is due to our sun being situated near the centre of an indefinite stratum of stars—the “disc theory,” as it is termed—was abandoned by its illustrious author in his later writings, and is now considered to be wholly untenable by nearly all astronomers who have studied the subject. Sir John Herschel remarks that the general aspect of the galaxy near the Southern Cross indicates “that the Milky Way, in this neighbourhood, at any rate, is really what it appears to be, a belt or zone of stars separated from us by a starless interval.” It certainly seems utterly improbable that the nearly circular blank space near the Southern Cross, known as “the coal sack,” should represent a tunnel through a disc, of which the thickness is comparatively small, while its diameter, on the “disc theory,” stretches out almost to infinity. A straight, tunnel-shaped opening of great length, pointing directly towards the earth, would form an extraordinary phenomenon even in a solitary instance; yet there are several somewhat similar openings to be found in the Milky Way, as viewed both with the naked eye and with a telescope. That all these openings should represent tunnels radiating from a common centre is quite beyond the bounds of probability, and, indeed, such an hypothesis does not deserve serious consideration. With reference to a photograph of the Milky Way in the constellation Cepheus, Professor Barnard says, “the sky (or Milky Way) is broken up into numerous black cracks or crevices. Looking at these peculiar features, I cannot well see how one can avoid the conclusion that they are necessarily real vacancies in the Milky Way, through which we look out into the blackness of space.”^[150] Using a telescope with a low power, Mr. S. M. Baird Gemmill says, “December 1, 1886. In sweeping over the constellation of Monoceros, I was much struck with the reticulated character of the arrangement of the brighter stars upon the glimmering background, and the way in which this background seemed to follow the reticulation. By ‘brighter stars’ are meant stars of from 8 to 10 magnitude, for it was among these that I noticed this peculiarity of arrangement. It put me in mind of M. M. Henry’s photographs of Cygnus. The region seemed, in fact, a vast network of stars, the reticulations of which were separated by desert, or comparatively desert spaces.”^[151] I have noticed the same thing myself while examining the Milky Way with a binocular field-glass. On October 26, 1889, I noted as follows: “North of Alpha Cygni, and near Xi and Nu Cygni, the nebulous light of the Milky Way seems to cling round and follow streams of small stars in a very remarkable way; numerous small ‘coal sacks’ and rifts are visible, in which comparatively few stars are to be seen with the binocular.” This observation has been fully confirmed by photographs of this region, taken by Dr. Max Wolf in 1891.

That the Milky Way is not indefinitely extended in the line of sight seems clearly shown by Sir John Herschel’s observations in the Southern Hemisphere. In his “Outlines of Astronomy” (p. 578), he says: “When examined with powerful telescopes, the constitution of this wonderful zone is found to be no less various than its aspect to the eye is irregular. In some regions, the stars of which it is wholly composed are scattered with remarkable uniformity over immense tracts, while in others the irregularity of their distribution is quite as striking, exhibiting a rapid succession of closely clustering rich patches, separated by comparatively poor intervals, and indeed, in some instances, by spaces absolutely dark and completely void of any star,^[152] even of the smallest telescopic magnitude.... In some, for instance, extremely minute stars, though never altogether wanting, occur in numbers so moderate, as to lead us irresistibly to the conclusion that, in those regions, we see fairly through the starry stratum, since it is impossible otherwise (supposing their light not intercepted), that the members of the smaller magnitude should not go on increasing ad infinitum. In such cases, moreover, the ground of the heavens, as seen between the stars, is for the most part perfectly dark, which again would not be the case if innumerable multitudes of stars, too minute to be individually discernible, existed beyond. In other regions we are presented with the phænomenon of an almost uniform degree of brightness of the individual stars, accompanied with a very even distribution of them over the ground of the heavens, both the larger and smaller magnitudes being strikingly deficient. In such cases it is equally impossible not to perceive that we are looking through a sheet of stars nearly of a size and of no great thickness compared with the distance which separates them from us. Were it otherwise, we should be driven to suppose the more distant stars uniformly the larger, so as to compensate by their greater intrinsic brightness for their greater distance, a supposition contrary to all probability. In others again, and that not unfrequently, we are presented with a double phænomenon of the same kind, viz., a tissue, as it were, of large stars spread over another of very small ones, the intermediate magnitude being wanting. The conclusion here seems equally evident that in such cases we look through two sidereal sheets separated by a starless interval.”

An examination of the evidence at present available, with reference to the distribution of the visible stars in space, has recently been undertaken by Professor Kapteyn of Groningen, and an account of the conclusions he has arrived at may prove of interest to the reader.