Popular Science Monthly/Volume 87/September 1915/The Evolution of the Stars and the Formation of the Earth
THE
POPULAR SCIENCE
MONTHLY
SEPTEMBER, 1915
| THE EVOLUTION OF THE STARS AND THE FORMATION OF THE EARTH[1] |
By WILLIAM WALLACE CAMPBELL
DIRECTOR OF THE LICK OBSERVATORY, UNIVERSITY OF CALIFORNIA
Introduction
EVERY serious student of nature asks, sooner or later: What was the origin of the stars? What has been their history? And what does the future hold in store for them?
In harmony with our experience is the belief that all matter in the universe is endowed with the property of obeying certain fundamental laws, such as: every particle of matter attracts every other particle; a hotter body radiates its heat energy to a cooler body; gases expand indefinitely unless resisted by gravitation or other effective force. Again, everything in nature is growing older and changing in condition; slowly or rapidly, depending upon circumstances; the meteorological elements and gravitation are tearing down the high places of the Earth; the eroded materials are transported to the bottoms of valleys, lakes and seas; and these results beget further consequences. In general, the changes in small bodies proceed rapidly and in great bodies slowly.
Astronomers believe there has been an orderly development of the stars, in obedience to precisely the same simple laws that govern our every-day affairs. Starting with the materials as already existing, our problem is to trace in outline the probable course of the evolutionary processes which have given us the stellar universe.
The effort to find a solution brings us against two superlative difficulties:
First, save only the Earth and an occasional meteorite, all the bodies that concern us are at tremendous distances. We must study them at long range, through the reading and interpretation of the messages which their own rays of light and heat carry from them to us. We bring bodies closer, in effect, by means of telescopes, but the reduced distances are still heroic. The stars, some of which are many millions of kilometers in diameter, are still seen as mere points of light in our most powerful telescopes, even though the telescopes magnify 3,000-fold.
Secondly, the evolutionary processes are exceedingly deliberate. We do not know that any progressive changes have ever been noted in any celestial body, except in the comets and meteorites, in the Earth's surface strata, and possibly in the so-called new stars. We observe changes in the clouds of Jupiter, changes in the surface features of the Sun, and some 4,000 stars are known to vary in brightness; but all these are short-period changes, and they do not indicate that progressive or permanent changes are involved.
We can get no help in our problem by waiting for any star to show signs of change in physical condition—we should probably have to wait tens of thousands, and perhaps millions, of years. We must take the heavenly bodies as they are, try to fit them into an orderly series representing the various stages of evolutionary development, and justify our arrangement by means of the evidence collected.
We need, first of all, to comprehend as thoroughly as possible what the individual heavenly bodies are, how they are arranged in space, and how they are related to each other, both physically and geometrically. At the cost of telling you many things you have already learned I shall recall a few features in the structure of the solar system and of the stellar system, and describe briefly the characteristics of each class of objects with which we have to deal.
The Solar System.
In the solar system we have the great central body, our Sun, around which revolve the 8 major planets and their 26 moons, the 800 minor planets or asteroids discovered to date, the zodiacal-light materials, the comets and the meteors. The Sun is one of the ordinary stars. It seems very large, very bright and very hot, because it is relatively near to us, and we receive from it our entire supply of energy; but, compared with the thousands of other stars visible on any clear night, it is merely an average star. Nevertheless, the Sun is a very large body; if it wore a hollow shell of its present diameter we could pour more than a million Earths into it and still leave empty the space between the earth-balls. Traveling outward from the Sun we come, first, to the small planet Mercury, its diameter a little more than one third the Earth's diameter, which revolves once around the Sun in 88 days; secondly, to the planet Venus, just a shade smaller than the Earth, with period of revolution 225 days; and thirdly, to the Earth and its moon, which revolve around the Sun in one year. Fifty per cent. farther out than the Earth is Mars, its diameter a trifle more than one half the Earth's, with two tiny moons, and period of revolution 1.9 years. Next
Fig. 1. The Zodiacal Light.
are the asteroids, about 800 discovered to date, which revolve around the Sun, each in its own orbit, in from 134 to 8 years, the orbits varying greatly in size, eccentricity and position of orbit planes; then we come to the giant Jupiter, its diameter 11 times the Earth's diameter, and 9 moons, the system completing a revolution about the Sun in 12 years; still farther out is Saturn, its diameter 9 times the Earth's, with its wonderful ring system and 9 moons, all revolving around the Sun in 2912 years; next is Uranus, 4 times the Earth in diameter, with 4 moons, all revolving around the Sun once in 84 years; and finally we come to the outermost-known planet, Neptune, a shade larger than Uranus, and its one moon, this planet requiring 165 years to travel around the Sun.
Again, as to the material which composes the solar system: its distribution is most remarkable. Nearly all of it is in the Sun. If we add together the masses of the major planets, the hundreds of asteroids, the satellites, make liberal allowance for the comets, etc., and call the total 1. then the mass of the Sun on the same scale is 744; that is, of 745 parts of matter composing our Solar System, 744 parts are in the Sun and only 1 part is in the bodies revolving around it. To state it differently, 9967% per cent. is in the Sun. and only 17 of 1 per cent. is divided up to make the planets, satellites, asteroids, comets and meteors. The four outer planets, Jupiter, Saturn, Uranus and Neptune contain 225 times as much material as the four inner planets, Mercury, Venus, Earth and Mars. The Earth is fully 3,000 times as massive as the 800 asteroids combined. There is the zodiacal-light material, which, in a more or less finely-divided state, as dust grains or very small bodies, revolves around the Sun, each separate particle in effect a minute planet. This matter, distributed through a great volume of space somewhat the shape of a double-convex lens, whose center coincides with the Sun, and whose edge extends out at least as far as the Earth's orbit, reflects and scatters the Sun's rays falling upon it. and causes the illumination easily visible after sunset in the west and before sunrise in the east. Then there are the comets which pass in orbits usually very elongated around the Sun, their tails pointing approximately away from the Sun; and the meteoric matter, which, at least in part, and quite possibly all, revolves around the Sun in elliptic orbits. Occasionally a meteorite gets through our atmosphere to the Earth's surface, is found and is installed in a museum; but many millions which collide with our atmosphere every 24 hours are consumed by frictional heat in the atmosphere and lose their identity.
It is a most remarkable fact that all the planets revolve in orbits lying nearly in the same plane. Let us call the distance from the Sun to the Earth 1; then the distance from the Sun to Neptune is 30; and the diameter of Neptune's orbit is 60. Now our system lies so nearly in one plane that we could put it in a very flat band-box 60 units in diameter and only 1 unit thick, so that all the major planets and their satellites, and all the asteroids with a very few exceptions, would perform their motions entirely within the box. The exceptional asteroids and the majority of the comets would dip out of the box because the planes in which their orbits lie make considerable angles with the central plane of the solar system.
It is an equally remarkable fact that the eight planets and the 800 asteroids are all revolving around the Sun in the same direction, which we call west to east. Likewise, the Sun rotates on its axis from west to east, and so also do Mercury, Venus, the Earth and its Moon, Mars, Jupiter and Saturn. Our moon, Mars's two moons, the seven inner moons of Jupiter, Saturn's rings and eight of its moons, revolve around their plants from west to east. From Jupiter out to Neptune we come upon exceptions to the rule. The eighth and ninth moons of Jupiter go around the planet from east to west. The ninth moon of Saturn is similarly reversed in direction. The four moons of Uranus move in a plane making an angle of 98° with the principal plane of the solar system; that is, nearly at right angles to the principal plane. The one moon of Neptune moves in a plane inclined 145° to the plane of the system; in effect, from the east toward the west. The equatorial planes of Uranus and Neptune are, without doubt, essentially coincident with their satellite planes.
The Stellar System.
Our solar system is very completely isolated from other systems. Light travels from our Sun out to Neptune in less than 414 hours, yet it requires 414 years to travel from our Sun to the nearest star, α Centauri. Stating the case differently, the nearest star is more than 9,000 times as far from our system as our farthest planet, Neptune, is from the Sun. We should have to go 7 light-years from our Sun in another direction to reach the second-nearest star. It is 9 light-years in a still different direction to Sirius. The average distances between neighboring stars, at least in our part of the universe, is 6 or 7 or 8 light-years. We can see that the stars themselves occupy very little space, and that they have an abundance of room to move about. Recalling, further, that the average speed of the stars is about 26 kilometers per second, which means that about 80,000 years would be required for the average star to travel over the average distance to its neighbor, we can see that collisions of two stars must be exceedingly rare; and that close approaches of two stars, approaches so close as to disturb each other violently, must also be rare. However, when we consider the number of stars in the stellar system, we should perhaps expect a few close approaches to occur in a human life time; possibly also a grazing collision, but probably no full collision.
The universe of stars—our stellar system—is believed by students of the subject, all but unanimously, to occupy a limited volume of space
Fig. 2. Milky Way in Constellation Cygnus near tee Star Gamma, photographed by Professor Barnard with the 10-inch Bruce camera of the Yerkes Observatory.
that is somewhat the shape of a very flat pocket-watch; more strictly, a much flattened ellipsoid or spheroid. However, it is not intended to convey the impression that the boundaries of the stellar system are sharply defined, or that the stars are uniformly distributed throughout the spheroid, and all at once, at the surface of the spheroid, cease to exist; but only that the stars are more or less irregularly distributed throughout a volume of space roughly spheroidal in form, and that the thinning out of stars near the confines of the system may be quite gradual and irregular. The equatorial plane of the spheroid is coincident with the central plane of the Milky Way. We see the Milky Way as a bright band encircling the sky, because in looking toward the Milky Way we are looking out through the greatest depth of stars. There is considerable uncertainty as to the dimensions of the system, chiefly for two reasons: first, the stars near the surface of the spheroid are everywhere too far away to let us measure their distances directly, and, in fact, so far away that we have not been able to measure their transverse motions—their proper motions—and thus to gain indirectly an idea of their distances; and secondly, the spheroid may be considerably larger than it seems because of possible, and even probable, absorption or obstruction of star-light in its passage through space. Newcomb has suggested that the shorter radius of the spheroid, at right angles to the plane of the Milky Way, may be taken as of the order of 3,000 light-years. The long radii of the spheroid, that is, the radii in the plane of the Milky Way, may be at least 10 times as great; that is, 30,000 light-years or more.
The solar system is believed to be somewhere near the center of the stellar system: the counts of stars in all parts of the sky indicate that the Milky Way structure is not much closer to us, so to speak, in one direction than in other directions; there are about as many stars on one side of a plane through the central line of the Milky Way as there are on the other. Wilhelm Struve's statistical studies of stellar distribution led him to conclude that the effective central line of the Milky Way is not a "great circle," but a "small circle," lying at a distance of 92° from the north pole of the galaxy and 88° from the south pole of the galaxy. Interpreted, this means that the solar system lies a short distance north of the central plane of the stellar system.
This conception of the stellar universe and Milky Way agrees in all important particulars with Immanuel Kant's ideas and description published in the year 1755—a remarkable contribution, based essentially on naked-eye observations, without the advantage of accurate observations laboriously made with telescopes. However, it was the star counts by the two Herschels, father and son, which put this conception of the stellar system upon the basis of confidence. Sir William Herschel, using an 18-inch reflecting telescope in the northern hemisphere, and Sir John Herschel. using the same telescope in the southern hemisphere, counted the stars visible in the eyepiece, 15 minutes of arc in diameter, in 7,300 regions distributed rather uniformly over the entire sky. They found that the number of stars decreased rapidly as they passed from the central plane of the Milky Way toward the north and south poles of the galaxy. Here is a table deduced by Struve from the Herschels' counts.
| Galactic Latitude Zones | Average Number of Stars Per Field '5' in Diameter | ||||
| +90°— +75° | 4.32 | ||||
| +75 — +60 | 5.42 | ||||
| +60 — +45 | 8.21 | ||||
| +45 — +30 | 13.61 | ||||
| +30 — +15 | 24.09 | ||||
| +15 —0 | 53.43 | ||||
| 0 — -15 | 59.06 | ||||
| -15 — -30 | 26.29 | ||||
| -30 — -45 | 13.49 | ||||
| -45 — -60 | 9.08 | ||||
| -60 — -75 | 6.62 | ||||
| -75 — -90 | 6.05 | ||||
The average number of stars in the Milky Way zone 30° wide, that is, in galactic latitude +15 to -15, visible in the eyepiece of the telescope, was 56, whereas in the region surrounding the north and south galactic poles the average number visible in the same eyepiece was but 5. The great condensation in the Milky Way is not fully evident from the table. The stars are much more numerous near the central line of the Milky Way than they are near its borders. The average number along the central line, found by Sir William Herschel, was 122. There is no reason to doubt that the preponderance of stars visible in the direction of the Milky Way is due to the greater extension of the stellar system in that direction than in the direction of the galactic poles.
It has been noted by several observers that the faintest stars visible in telescopes of moderate size, that is, stars of the 14th, 15th and 16th magnitudes, are plentiful in the Milky Way and very scarce at a distance from the Milky Way. The contrast between Milky Way and non-Milky Way regions is scarcely noticeable in the naked eye stars, but it becomes stronger and stronger as we pass to the fainter stars.[2]
If there is an absorption of light in its passage through space, such that the very distant stars are appreciably reduced in brightness, then the stars of average size and physical condition must be invisible to us when they are farther away than a certain limiting distance, and in that case the extent of the universe in the direction of the Milky Way may be vastly greater than we have described it; but this consideration would not act to increase the radius of the actual stellar system in the direction of the poles of the galaxy by any appreciable amount.
Investigations conducted principally at the Harvard and Greenwich Observatories indicate that the number of stars visible in our largest telescopes is of the order of 60,000,000 or 70,000,000, and that the number which can be recorded on photographic plates by means of long exposures with our largest reflecting telescopes is several times as great.
Investigations by Newcomb and Kelvin upon the gravitational power of the stellar universe to produce the observed velocities of the stars give indications that the visible stars contain in reality only a fraction, perhaps one fifth, of the gravitating materials concerned, and they conclude that more material exists in dark and invisible stars than in the visible ones. I am inclined to regard their estimates of dark material as of questionable accuracy, on account of the purely arbitrary assumptions involved.
Stellar Motions
It is necessary that we consider briefly the motions of the stars, including that of our own star. It has been found that all celestial bodies, as far as they have been studied, are in motion with reference to the entire system, and with reference to each other. Our Sun is no exception to the rule: it is traveling rapidly through the stellar system, carrying its planets and their satellites along with it. The apparent motions of the individual stars are not in general their real motions: they are a compound of the real motions and of our motion. If the other stars were really at rest in the great system, they would still seem to be moving because our star is carrying us past them, so to speak: the nearer stars would seem to be moving rapidly, and the more distant stars less rapidly, away from that point in the sky which we are approaching. Since the stars are really moving in a great variety of directions, with a great variety of speeds, their apparent motions are also in a great variety of directions, but the prevailing tendency of their motions is away from our goal.
By studying these compounded motions, Herschel, in 1783, and a long line of distinguished investigators following Herschel, have established that our solar system is traveling toward a point in distant space about on the boundary line between the constellations Hercules and Lyra. If the solar system is moving rapidly toward that point, the stars in that vicinity should, on the average, seem to be approaching us, and the stars in the opposite region of the sky should, on the average, seem to be receding from us. The spectrograph enables us to measure the rates of approach and recession of the individual stars. It has been found that while the hundreds of bright stars in the Hercules-Lyra region are traveling, some away from us and some toward us, with a very great variety of speeds, yet, on the average, that group of stars seems to be approaching ns at the rate of 19 kilometers per second. In a similar manner it has been found that the stars near the opposite point of the sky, while moving individually with a great variety of velocities of approach and recession, are, on the average, receding from the solar system with a speed of 19 kilometers per second. No one questions the explanation of these facts: the solar system is traveling toward the Hercules-Lyra region with a speed of 19 kilometers per second. If, now, the speed of 19 kilometers per second be maintained, and the longer radii of our stellar system be 30,000 light years, we should require a period of 450,000,000 years to travel from the center to the circumference of our system. The youth of the solar system was probably spent in a very different part of the stellar system from where it now is.
Comets
Are the comets bona fide members of the solar system as the planets are, or are they transient visitors from the greater stellar system? Immanuel Kant in 1755 advocated the view that the comets are genuine members of the solar system. From 40 to 70 years later Laplace advocated the other view, that the comets belong to the great stellar system, and that a few of them happen, in the course of their travels, to encounter the solar system. The latter view prevailed from Laplace's time almost up to to-day. If the comets are of our solar system they should move in elliptic orbits; that is, they should return again and again to the vicinity of the Sun.
If the Sun were at rest with reference to the stellar system and the comets should start with exceedingly small velocities from a very great distance, say 20 or more light years away, they would travel around our Sun in curves which we could not distinguish from parabolas. Interpreted, this means that they would eventually go back to approximately the same distant region of space from which they started and never again return to the solar system. If the comets should start toward us, from interstellar space, with appreciable velocities, they would move around the Sun in hyperbolic orbits, curves whose branches, one coming in toward the Sun and one going out from the Sun, diverge widely; such comets would go away to a region of space totally different from that which they had occupied before their solar visits and never return either to us or to their original habitation. Since the Sun is not at rest in the stellar system, but is traveling 19 kilometers per second toward the Lyra-Hercules constellations, it can be shown that the forms of the orbits of comets coming from interstellar space, whether they start from rest or with the average speed of the stars, would, in general, be strongly hyperbolic. The observed facts are that of the more than 400 cometary orbits determined, only 8 or 9 have been suspected to be hyperbolic. Further, the recent researches of Fabry and Strömgren have shown that all of the suspected cases either rest upon insufficient observations of the comets at the time of their appearance, so that the orbits are uncertain, or that the disturbing attractions of our planets have converted the orbits from the elliptic to the hyperbolic form after the comets have got well within our planetary system. Another fact is equally important. By virtue of our rapid travels toward the Lyra-Hercules region we should meet more comets coming from that direction than there are comets overtaking us from the opposite direction. To state this point differently: of the comets which swing around the Sun, a greater number should have come into our system from the Lyra-Hercules region than from any other region, and especially from the region of sky which we are leaving behind. The facts are otherwise: we can not say that the approaches of comets favor any particular direction.
The orbits of the great majority of comets are very close to the parabolic form. The nature of comets is such that they are under observation for a few weeks or a few months, and only an occasional one for a year or more. When but a small section of the orbit has been thus observed it is difficult to decide between the parabola and a very elongated ellipse. It happens, however, when these comets have been accurately observed through many months, and the disturbing attractions of our planets have been taken into account, that the orbits are found to be very long ellipses and not parabolas; some of the ellipses are so elongated that thousands, and occasionally hundreds of thousands of years, are required to complete one circuit of the Sun. Let us assume that a comet belonging to the solar system starts at rest, with reference to the solar system, from a point midway between α Centauri and our Sun, and travels around our Sun. It would be 60,000,000 years in reaching us, or 120,000,000 years in completing its circuit. It is evident that an immense amount of cometary material must exist in the outer regions of our Sun's gravitational field in order that a minute part of it may visit the Sun every three months, which is about the average interval of time between the coming of these bodies.
It should be noted that the planes of the very elongated comet orbits show no preference for small angles with the plane of the solar system: they intersect the solar system plane at all angles, and these comets come into our system from all directions indifferently.
We must hold, I think, that the comets are genuine members of our solar system: the great majority spend most of their time in the outer parts of our system, far beyond the orbits of Neptune, but they are moving through space as companions to our Sun as truly as the Earth and Jupiter are.
Aside from the comets which come from great distances, there are, of course, the so-called periodic comets which move in relatively short ellipses, revolving around the Sun in a few years, and reappearing at predicted times and places. About 60 such comets have been observed with periods less than 100 years. There is the great Jupiter family of comets, about 30 in the family, so-called because their aphelions—the points of the orbits farthest from the Sun—lie near Jupiter's orbit. Their periods vary from 3 to 8 years, their motions around the Sun are all from west to east, as in the case of the planets, and their orbit planes make small angles with the plane of the solar system. In a similar way there are two Saturn comets, three Uranus comets, and six Neptune comets, one of the latter being Halley's. Halley's comet is revolving around the Sun from east to west; that is, in a retrograde direction; and the motions of two comets which disappeared many years ago were likewise from east to west. The motions of all the other short-period comets are from west to east.
The origin of the periodic comets is an interesting question. Newton, of Yale, who was the chief student of the subject, gave practical certainty to the view that the periodic comets have been captured, so to speak, by the major planets, and especially by Jupiter; that is, that comets approaching the Sun in their elongated orbits and passing close to the major planets have had their orbits converted, either during one visit or cumulatively during several visits, into the forms we now observe. Perhaps the strongest doubt as to the sufficiency of the explanation arises from the fact that 95 per cent. of the motions appear to be from west to east. Newton's theory seems to demand that about 25 per cent. of Jupiter's comets should move in retrograde orbits, whereas none of Jupiter's comets, nor the two Saturn comets, do so move. Three of the eleven comets related to Uranus and Neptune, namely, Halley's comet and two lost comets, travel in the retrograde direction. The capture theory is technical and we must not pursue it. Fortunately, there is another avenue of approach. Barnard has noted that the short-period comets differ in appearance from those which come to our system unexpectedly, in that the former are the more diffuse in appearance; that is, they have larger diameters in proportion to their total brightness. There is reason to believe that the head of a comet consists principally of separate small bodies. Now in a collection of small bodies the gravitational forces holding them together are extremely slight. When the group approaches the center of the solar system the Sun's attractions upon the nearer members of the group are appreciably stronger than upon the members which are farthest from the Sun. The orbital motions of the nearer particles are relatively quickened and those of the farther particles relatively delayed. If the comet is traveling upon a very elongated orbit the mutual attractions within the head can again be effective while the comet is in the outer parts of the orbit, and a condensing process probably occurs; but, if the orbit extends out only as far as Jupiter or other major planets, there is little opportunity for the internal attractions to re-condense the particles, and the next approach to the Sun carries the scattering process a step further. Repeated returns to the Sun dissipate the individual constituents of the comet more widely. The intensity of the comet's light is reduced, and eventually it becomes too faint for discovery and observation. There is little room to doubt that this process is responsible for the total disappearance of several periodic comets.
Meteor Streams
The argument is strongly supported by the meteor streams. It is well known that on certain nights of the year we see an unusually large number of meteors, which come from certain definite directions in space. These meteors have been extensively observed and their orbits have been computed. The illustration shows the orbits of four such swarms.
Fig. 3. Orbits of Meteoric Swarms, which are known to be associated with comets.
They intersect the Earth's orbit at certain computed points. We pass through those points on certain nights of the year and the meteoric materials moving in the one orbit collide with the Earth in the other orbit. Now, it has been shown that the orbits of these four meteor streams and of one other stream are the orbits of five periodic comets which have disappeared from sight. Clearly, the cometary materials had been gradually scattered by the disintegrating effect of the Sun's attraction, and the separate particles were compelled to move in orbits differing slightly from each other, and from the recognized orbits of the comets. The meteoric collisions with the Earth are such as to show that we are dealing with widely separated small masses moving in orbits nearly identical with each other.
In the case of these five swarms there is certainly a close connection between meteors and comets. Whether all meteoric matter has come from the disintegration of comets can not be answered now. We can say that since the Earth actually passes through at least five prominent meteor swarms,[3] there ought to be thousands of invisible swarms within our solar system which we do not pass through. Newton's investigations led him to the conclusion that about 90 per cent. of the meteors which have encountered the Earth and have been observed with sufficient accuracy to let us determine their orbits are moving around the Sun in eccentric orbits of short periods, like those of the short-period comets, and in the west-to-east direction.
The certainty of rapid disintegration of the periodic comets—extremely rapid in comparison with astronomical time-intervals—is all but equivalent to saying that the periodic comets have been recently captured by our planets; for the periodic comets which we are still observing could not have been following their present orbits during many centuries, except at the price of disintegration to the point of total disappearance.
The Zodiacal Light
The zodiacal light is a closely related subject. The phenomenon is due to the presence of countless small particles of solid matter varying perhaps from dust particles up to bodies perhaps many cubic inches or even larger in volume, which scatter the sunlight falling upon them. The volume of space occupied by this finely divided material is very great. It extends north from the Sun to a distance of the order of 100,000,000 kilometers, and there is no reason to doubt an equal southern extension, for observations made in the west after sunset and in the east before sunrise indicate that the structure is symmetrical with respect to the Sun. Its extent in the principal plane of the solar system in all directions from the Sun is even greater. In such clear skies as exist on the tops of mountains the zodiacal light can be seen to stretch entirely across the sky as a faint band following the ecliptic; and this is proof abundant that the materials which scatter the light extend beyond the Earth's orbit.
Inasmuch as we do not distinguish the individual particles which make up the zodiacal light materials, we can not now say whether they are revolving around the Sun from west to east, but we can not doubt the fact that they are revolving around the Sun and that the orbits of a large proportion of the particles are necessarily in planes highly inclined to the general plane of the solar system. Seeliger is of the opinion that this material supplies the attracting mass which disturbs the motion of Mercury, and to a lesser degree the motions of Venus, Earth and Mars. If this be true the total mass of the particles must approximate to that of the planet Mercury.
Where this material came from, whether it is a remnant of the original material which formed the inner planets and the Sun, or whether it has come in from the outer confines of the Sun's sphere of influence in the same way that the comets have transported very distant materials into the terrestrial region, is wholly unknown.
Stars Clusters
The star clusters offer a wide range of character, as to their density of stellar contents and as to the symmetry of distribution. There are the large irregular clusters visible to the eye, such as the Pleiades, Praesepe, the Perseus clusters, in which the stars are widely separated and irregularly distributed. There are the globular clusters, invisible to the naked eye, except in three or four cases, which contain multitudes of faint stars densely crowded together and quite symmetrically arranged. The great cluster in Hercules is the most striking example in the northern skies. The accompanying photograph, secured with the 60-inch reflector of the Mount Wilson Observatory, records stars to the order of 30,000, each star a sun as truly as is our star. There are two still more extensive clusters in the Southern Hemisphere, but they have not yet been photographed on the same scale as the northern clusters. The globular clusters, of various degrees of stellar richness, exist to the number of several scores.
There are two great agglomerations of stars—two dense clouds of stars—occupying isolated positions in the far southern sky, quite distant from the Milky Way, which seem to have many of the Milky Way's attributes. They appear to be great irregular clusters of stars differing only in size from the vastly greater Milky Way cluster. These objects are known as the Greater and Lesser Magellanic Clouds.
The Nebulae
The objects which probably concern our subject most directly are the nebulae. The word nebula means a "little cloud"; and like little clouds superimposed upon the dark background of the sky the first 10,000 nebulae looked to their discoverers. They were of various sizes, from that of the Orion nebula, and even larger, down to those indistinguishable in small telescopes from stars, and to those so faint as to be on the limit of telescopic vision. The Herschels, father and son, were the first great discoverers of the nebulæ. Lord Ross's reflecting telescope showed that a few of the very bright and large nebulæ, perhaps two dozen in all, are not formless masses, but spirals—indicating plainly that they have motions of rotation. It was noticed by Sir William Herschel, a century ago, that the distribution of the nebulæ on the surface of the sky is most remarkable. Proctor's chart, published in 1869, illustrates this fact. On this chart the cloud-like forms of the Milky Way are outlined across both hemispheres, as seen by the naked eye, but it should be said that telescopic vision of the Milky Way would present very different and vastly more uniform outlines. Each dot on the chart represents a nebula. He who runs may
Fig. 4. The Great Star Cluster in Hercules. Photographed at the Mount Wilson Solar Observatory.
read that the nebulæ in general abhor the Milky Way. In the northern hemisphere they cluster most densely in the neighborhood of the pole of the galaxy. In the southern hemisphere they show the same tendency, but not so strongly. There are nebulae in the Milky Way, but
| Southern Hemisphere. | Northern Hemisphere. |
| Fig. 5. Distribution of Nebulæ (and Star Clusters). According to Proctor. Nebulæ are marked by dots; clusters by crosses. | |
they are relatively few. Herschel's and Proctor's conclusions related only to the brighter nebula?, which had been discovered by visual methods.
Before Keeler began to photograph nebulæ with the Crossley reflector, in 1898, some 10,000 of these bodies had been discovered and catalogued. A few plates exposed by Keeler here and there over the northern sky recorded several hundred additional nebulæ. Using his photographs of small areas of the sky as samples, he estimated conservatively that at least 120,000 nebulæ are discoverable with the Crossley reflector. Further observations by Perrine with the same instrument and by Fath with the 60-inch Mount Wilson reflector have shown that the number discoverable with fairly short exposures is considerably greater than 120,000. Path's plates, uniformly distributed over the northern sky from the North Pole to declination 22°.5 south of the Equator, recorded 864 nebulæ previously unseen. The numbers on the individual plates are set down in the corresponding area. The curve drawn across the chart represents the central line of the Milky Way. The north pole of the galaxy is at N. The distribution of these faint nebulæ is seen to be patchy, but the fact is in evidence that the faint nebulæ, like those bright enough to be discovered by visual methods, abhor the Milky Way. Keeler's photography of the nebulæ led him to open another chapter in nebular investigation with the startling discovery that "most of the nebulæ have the spiral structure." This applied not only to the faint nebulæ which he discovered, but to the nebulæ already known. Keeler's successors have confirmed this discovery: it is certain that the great majority of the nebulæ have the spiral form. What the relative number of spirals and formless nebulæ may be remains for the future to decide.
The spirals vary all the way from the great. Andromeda nebula down to those so small that the photographic plate is just able to separate the details of structure; and there is no reason to doubt that more powerful instruments would show still smaller objects to have the spiral structure.
There are irregular nebulæ of all sizes. The brilliant Orion nebula is diminutive in size compared with the faint nebulosity, discovered by William H. Pickering in 1889, which forms the background for almost the whole of the constellation of Orion. The well-known nebulous structure connected with the brighter Pleiades stars is small in comparison with the area covered by a faint exterior nebulosity discovered by Barnard in 1893. There are very great irregular nebulæ such as the Network Nebula in Cygnus and the nebulous background in the Greater Magellanic Cloud. Barnard's wonderful photographs of the Milky Way have recorded many extensive nebulous fields, especially in regions where the background of the galaxy shows relatively few stars (see Fig. 2).
The so-called planetary nebulæ are of special interest, as we shall learn in the sequel. Small in size, all more or less dense, some quite 
Fig. 7a. Planetary Nebula, N. G. C. 2.392. Photographed at the Lick Observatory.
Fig. 7b. Planetary Nebula, N. G. C. 40. Photographed at the Lick Observatory.
Fig. 7c. Planetary Nebula, N. G. C. 7.009. Photographed at the Lick Observatory. Two exposures, one short and one long.
regular in outline, and a large proportion containing condensed or stellar nuclei near their centers, they were called planetaries by Herschel, because, though faint, they present discs somewhat as the planets do, when viewed under low power.
There are several scores of so-called stellar nebulæ. In moderate-sized telescopes most of them look like ordinary stars. In large telescopes many of them are hazy, but some are as well defined as stars. The spectroscope shows that all are true nebulæ. If they were much closer to us we should doubtless see them as planetary nebulæ.
A few other interesting objects are known as ring nebulæ, the most noted case being the ring nebula in Lyra.
Among the remarkable facts of the stellar universe are these: the large irregular nebulæ, the ring nebulæ, the planetary nebulæ, and the stellar[4] nebulæ, with relatively rare exceptions, are in or very close to the Milky Way: and, on the contrary, the spirals in or near the Milky Way are of negligible number. The first group are without question an integral part of our stellar system. The spirals, seem not to be closely connected with our stellar system, yet their very avoidance of the Milky Way shows that they bear some intimate relationship to it. There is no occasion for surprise that a small group of special objects should be in the Milky Way structure; but that the scores of thousands, and perhaps hundreds of thousands, of spirals, should abhor the Milky Way is a fact which immediately arrests our attention and calls for explanation. Moore has suggested that their absence from the Milky Way may be apparent and not real: that any absorbing or obstructing medium in the Milky Way structure might prevent the light of the spirals from reaching us, especially if the spirals are extremely distant. If the light from very distant nebulæ is absorbed or obstructed, as a function of the angular distance from the galaxy, the nebulæ near the poles of the galaxy, other things being equal, should on the average be intrinsically brighter than the nebulæ in or near the Milky Way. Secondly, if such an effect exists, long-exposure photographs on regions near the galaxy should record nebulæ in numbers more nearly equal to those recorded by short exposures near the poles of the galaxy. An examination of existing Crossley reflector photographs has led to negative results on this question, and we must assume that the spiral nebulæ really avoid the Milky Way.
The question of the distances of the spiral nebulæ has long been held in mind. The evidence, to which we shall refer later, is to the effect that they are very far away, and accordingly that they are of enormously
Fig. 8. The Ring Nebula in Lyra. Photographed at the Lick Observatory.
Fig. 9. Spiral Nebula in Canum Venaticorum, M. 51. Photographed at the Lick Observatory.
Fig. 10. Spiral Nebula in Leo. M. 65. Photographed at the Lick Observatory.
great dimensions. This is the particular reason why a few astronomers suggest that the spirals may be distant systems of stars. They say that our own stellar system, if viewed from a great distance, might be seen to have a spiral structure: 
Fig. 11. The Spiral Nebula H. V. 41, Canum Venaticorum, seen edgewise, Photographed at the Lick Observatory. that it would be fairly circular in general outline if viewed from the poles of the Milky Way, or greatly elongated or spindle-shaped if the observer were in the plane of the Milky Way. We illustrate this point by means of well known spirals viewed broadside, and obliquely, and of the spindle-shaped nebulæ, which we do not doubt are spirals seen edgewise. Easton, the principal modern student of Milky Way structure, has even gone through the laborious task of assigning the stars, as seen from our viewpoint near the supposed center of the stellar system, to their assumed places in a spiral structure. But I need scarcely say that the subject is too vast for solution now, or in the near future. If our stellar system is one of a hundred thousand or more spiral nebulæ, we have at once the problem of determining the place of our stellar system in the larger universe of systems, necessarily beginning with the motion of our system as a whole with reference to the great numbers of surrounding systems.
Star Streams
It was supposed, until ten years ago, that the stars are moving approximately at random, both as to direction and as to speed. In 1904 Kapteyn announced, on the contrary, that the stars have decided preferences for motions toward two opposite points in the sky; one point in the northern edge of Orion, in the Milky Way; and the other point exactly opposite to this. Investigations by many others have in all cases confirmed Kapteyn's discovery. Kapteyn did not mean to say that the individual stars are moving parallel to a straight line joining these two opposite points, but simply that their components of motion parallel to this line are considerably greater, on the average, than the components in any other direction. We may visualize his ideas in the following manner:
Assume the existence, ages ago, of a great cluster or cloud of stars distributed more or less uniformly through a certain vast volume of space, whose individual motions were at random in both magnitude and direction. Assume the existence of an entirely similar group of stars, occupying another vast volume of space, whose internal motions were also at random. Assume, further, that these two groups of stars were traveling through space in such a way that they more or less completely interpenetrated, with the result that the two groups of stars have now become a single group. There are stars still moving in all directions, with speeds of all dimensions within certain limits, and yet there exists a preference for motion along and parallel to the line which originally joined the centers of the two groups. Assume now that our Sun is carrying the terrestrial observer through the combined group in a direction making a considerable angle with the line of preferential motion: the apparent motions of the individual stars, as observed from the solar system, would then have preferences for two directions very different from the line joining the two original positions of the groups; we should find a great number deviating by small angles from the two preferential directions, a small number deviating to a greater extent, and relatively few whose motions make large angles with the preferential directions; and this is as the apparent motions of the stars have been determined by many observers.
Kapteyn's results depend upon proper motion data; that is, upon their apparent motions on the surface of the sky. Spectroscopic observations of stellar motions of approach and recession confirm that the stars have preferential motions, but to a smaller degree than proper motion data had indicated.
No one doubts that preferential motions exist, but the explanation is another matter. Kapteyn does not insist that our stellar system has actually resulted from the intermingling of two star streams, yet he inclines more and more to this point of view, and the hypothesis does seem to accord better with the observed facts than any other hitherto proposed. A strong objection to it is its apparent improbability. It does not seem reasonable that two great clouds of stars, containing all the stars now in our sidereal universe, should have come together and interpenetrated so completely as to have produced in an age when we happen to be the observers a stellar system apparently spheroidal in form. When I look at the Milky Way, completely encircling the sky, my mind is filled with doubt. And if two great galaxies of stars have traveled far and come together, they will travel further and through each other and we shall have two galaxies again, moving away from each other. It does not follow that the more distant and fainter stars will show the same preferential motions as the brighter and nearer ones which led Kapteyn to his hypothesis, though it should be said that a fairly extensive study of stars fainter than the Kapteyn stars made by Comstock led to results in good agreement with Kapteyn's. May it not be possible that the preferential motions observed are in some way connected with rotational phenomena within our stellar system, especially as the line of preferential motions lies approximately in the plane of the Milky Way, or are local to what we may call our region of the system, and not be true of the system as a whole?[5]
An alternative hypothesis of prevailing stellar motions, proposed by Schwarzschild, seems to have advantages from the point of view of probability, but it appears not to accord so well with the facts of observation. Schwarzschild suggests that if from a given point we draw vectors whose directions and lengths represent the directions and speeds of existing stellar motions, then the outer extremities of these vectors will define the surface of an ellipsoid (of preferential motions) having three unequal axes.
(To be continued)
- ↑ Second course of lectures on the William Ellery Hale Foundation, National Academy of Sciences, delivered at the meeting of the Academy in the University of Chicago, on December 7 and 8, 1914.
- ↑ A recent study of Mr. Franklin Adams's excellent photographs of the sky, by Messrs. Chapman and Melotte, shows a considerably smaller disparity in the numbers of faint stars in the galactic and non-galactic regions than the Herschels and others found.
- ↑ Numerous minor streams are reported by meteor observers.
- ↑ The terms irregular, ring, planetary and stellar are intended merely to differentiate these objects as to their appearance in the telescope or on the photographic plate. They do not in themselves indicate differences in constitution or physical condition. The ring, planetary and stellar nebulae have a great many characteristics in common.
- ↑ Turner has proposed the following explanation of the two star streams: The whole mass of the stellar system exerts a gravitational influence on the motion of each star in the system, and the individual stars revolve around the center of mass of the system in their elongated orbits. One star stream comprises all those stars moving away from the center, and the other stream all those stars moving toward the center. We can not doubt that the motions of the individual stars are influenced by the gravitational attractions of the stellar system and of the group of stars nearest to them; but observational data on stellar motions must be vastly more extensive than at present in order to test Turner's hypothesis.
Halm, of the Cape of Good Hope, has given evidence of the existence of a third star stream, much less extensive than Kapteyn's two streams.