WORLDS-OP f T . CORE- LIBRARY OF THE UNIVERSITY OF CALIFORNIA. Class THE WORLDS OF SPACE THE SPIRAL NEBULA 51 MESSIER IN CANES VENATICI. P. 213.) THE WORLDS OF SPACE A SERIES OF POPULAR ARTICLES ON ASTRONOMICAL SUBJECTS. BY J. E. GORE, F.R.A.S., M.R.I.A., ETC., HONORARY MEMBER LIVERPOOL ASTRONOMICAL SOCIETY, CORRESPONDING MEMBER OF THE ASTRONOMICAL AND PHYSICAL SOCIETY OF TORONTO; AUTHOR OF "PLANETARY AND STELLAR STUDIES," "THE SCENERY OF THE HEAVENS," "THE VISIBLE UNIVERSE," ETC. " In fields of air he writes his name, And treads the chambers of the sky ; He reads the stars, and grasps the flame That quivers in the realms on high." SPRAGUE. A. D. INNES AND CO., BEDFORD STREET. 1894. GENERAL RICHARD CLAY & SONS, LIMITED, LONDON & BUNGAY. PREFACE. THE following series of short popular articles on Astronomy and kindred subjects have been published in various magazines during the last three years, and deal chiefly with recent discoveries and advances in the science. My thanks are due to the Editors and Publishers of The Parents' Review, The Monthly Packet, The Newbery House Magazine, Knowledge, The Gentleman's Magazine, The Sun, The Arts Monthly, and Indian Engineering, for permission to reprint them, and to Dr. Roberts for some of his beautiful photo- graphs. I have also to thank Mr. Ranyard for permission to reproduce some of the photographs which have appeared in Knowledge. Chapters xxix., xxx., xxxi., and xxxiii. originally appeared in Indian Engineering, a weekly Journal published in Calcutta by Pat. Doyle, C.E., F.R.A.S., etc. J. E. G. February 1894. Q-I CONTENTS. CHAP. PACK I. ARE THE PLANETS HABITABLE? ... ... I II. TERRESTRIAL AND SUN-LIKE PLANETS ... 1 6 III. LIFE IN OTHER WORLDS... ... ... 3! IV. THE RELATIVE BRIGHTNESS OF THE PLANETS 41 V. A DOUBLE PLANET ... ... ... $2 VI. ALPHA CENTAURI AND THE DISTANCES OF THE STARS ... ... ... ... 57 VII. THE SUN AMONG HIS PEERS ... ... 68 VIII. REVOLVING SUNS ... ... ... 8 1 IX. WEIGHING THE STARS ... ... ... 95 X. THE MASS AND BRIGHTNESS OF BINARY STARS 105 XI. THE ORIGIN OF DOUBLE STARS ... ... 114 XII. VARIABLE DOUBLE STARS ... ... 123 XIII. SOME PECULIARITIES OF THE VARIABLE STARS 133 XIV. THE " DEMON " STAR ... ... ... 141 XV. COLOURED STARS ... ... ... 151 XVI. SIRIUS AND ITS SYSTEM... ... ... l6l XVII. DARKENINGS OF THE SUN ... ... 17 1 XVIII. THE .. R AND DISTANCE OF THE VISIBLE STARS ... ... ... ... 175 XIX. SWARMS OF SUNS ... ... ... 185 XX. GREAT NEBULAE ... ... ... 197 viii CONTENTS. CHAP. XXI. SPIRAL NEBULAE XXII. PLANETARY NEBULAE ... XXIII. CELESTIAL PHOTOGRAPHY XXIV. ASTRONOMY WITHOUT A TELESCOPE XXV. RECENT ADVANCES IN ASTRONOMY XXVI. SOME ASTRONOMICAL ERRORS AND ILLUSIONS XXVII. THE ARITHMETIC OF ASTRONOMY XXVIII. THE MYTHOLOGY OF THE STARS AND PLANETS ... XXIX. THE FIGURE OF THE EARTH XXX. THE VARIATION OF LATITUDE XXXI. THUNDERSTORMS AND AURORAS ... XXXII. THE BAROMETRIC MEASUREMENT OF HEIGHTS XXXIII. THE OBSERVATORY ON MONT BLANC 25 26', 28l 289 3 OI 314 320 325 335 THE WORLDS OF SPACE. i. ARE THE PLANETS HABITABLE? I DO not ask, Are the planets inhabited ? That is a question to which an answer will probably never be vouchsafed to man. Even in the case of our nearest celestial neighbour, the Moon, the highest powers of our largest telescopes would fail to reveal the existence of any living creatures on its surface. Animals the size of elephants, or even as large as the gigantic saurians of geological times, would be quite invisible in the giant telescope of the Lick Observatory. Large cities, or buildings of great extent if they existed might possibly be discerned, and thus afford evidence of the existence of intelligent beings. But this only on the Moon. All the planets are much too distant to enable us to see anything but markings on their discs, these markings being only dimly visible on most of them, and possibly denoting, in some 2 THE WORLDS OF SPACE. cases, the existence of oceans or large tracts of land on their surface. I say possibly, for only in one case that of Mars are these markings clearly enough defined, and sufficiently persistent in their character, to render it probable that they represent land and water. In the case of the Moon no definite indica- tions of life on its surface have ever been detected. Indeed, the known absence of air and water is evidence enough to prove that life is impossible on the surface of our satellite. We must, therefore, abandon the hope that any improvement in our telescopes will ever show indications of life on the planets. There is nothing, however, to prevent us from making an inquiry into the possibility of the planets being habit- able by living creatures such as we are familiar with. We may consider their position in the solar system with reference to the Sun, and their probable physical condition, so far as observation enables us to judge. We may in this way arrive at some conclusion, from analogy, as to the possible habitability of the planets and their satellites. The ancient philosophers thought that the Sun itself might possibly be inhabited ! Even in modern times this idea has been revived. Dr. Elliott in 1787 upheld this view, and on his trial at the Old Bailey for the murder of Miss Boydell, his friends maintained his insanity, and quoted as proof of their assertion the pages of his book, in which this opinion was expressed. Such an hypothesis as the habitability ARE THE PLANETS HABITABLE ? 3 of the Sun does not seem worthy of serious consideration. With reference to the planets, let us take them in order of their distance from the Sun, commencing with Mercury, the nearest to the solar orb. The mean distance of Mercury from the Sun, compared with that of the Earth, is about as 36 to 93, or 1 2 to 31. Hence, as light and heat vary inversely as the square of the distance, the average intensity of the Sun's light and heat on Mercury exceeds that on the Earth in the proportion of the square of 31 to the square of 12, or about 6f times. Owing, however, to the elliptical shape of Mercury's orbit, its distance from the Sun varies in the ratio of 43 to 29. The intensity of the solar heat will therefore vary during the course of its short year about 88 of our days in the proportion of the square of 43 to the square of 29, or about as 9 to 4. This violent change of temperature, which takes place in the short period of six weeks (half the period of revolution) is, of course, strong evidence against the hypothesis of any life existing on the surface of Mercury. Certainly none of the larger animals that we are familiar with could possibly withstand a heat of nine times the intensity experienced in the Earth's equatorial regions. The period of rotation of the planet on its axis, or the length of its days, has always been a matter of much uncertainty. Owing to the position of its orbit its distance from the Earth is subject to great 4 THE WORLDS OF SPACE. fluctuations. When near "inferior" conjunction, or on this side of the Sun, it shines at its brightest, but in this position it only appears in the form of a crescent, like the Moon when a " few days old." When near "superior" conjunction, or beyond the Sun, it shines with nearly a full face, but then the diameter of its disc is so small owing to its great distance from the Earth that any markings which may exist on its surface are only seen with much uncertainty. The few occasions on which Mercury is visible in the morning or evening twilight also render observations of its surface markings a matter of great difficulty. From observations of some spots on its surface, made by the German astronomer Schroter towards the close of the eighteenth century, Bessel concluded that the planet rotated in about twenty-four hours and fifty-three seconds, on an axis inclined about 70 to the plane of the planet's orbit. If this were the case the seasons on Mercury would not differ much except in their length from those of the Earth, and possibly near the poles, the planet's " arctic " regions, the temperature might be sufficiently cool to admit of some forms, at least, of animal life. The near coincidence of the period found by Schroter and Bessel with the length of our day looks, however, suspicious. If Schroter observed Mercury on several consecutive evenings (or mornings) at nearly the same hour, as he probably did, and watched the same spot ARE THE PLANETS HABITABLE? 5 on its surface, it is easy to see that even if the planet had no rotation on its axis he might have concluded, with some show of plausibility, that the period of rotation was about twenty-four hours. He might, of course, have concluded also that the rotation took place in some aliquot part of twenty- four hours, say, twelve hours, or six hours, but this, from the analogy of the other " terrestrial" planets (Venus, the Earth, and Mars), seemed an improbable hypothesis. In later years further observations of Mercury were made by the astronomers Biilow and Dr. L. de Ball, but their results seem to have still left the matter an open question. The problem was again attacked in 1881 by the famous Italian astronomer Schiaparelli, who, recog- nizing that the uncertainty of the results derived from the earlier observations was in a great measure due to the unfavourable visibility of Mercury in the morning or evening sky, resolved to observe the planet by an entirely new method. This new de- parture consisted in observing the planet in the day- time by means of an 8-inch refractor, and (latterly) an 1 8-inch refractor. Of course the exact position of the planet in the sky at any time can be easily com- puted, and the telescope pointed to its exact place. From the year 1881 to November 1889, Schiaparelli succeeded in observing Mercury on 150 days. On one occasion he observed the planet when distant from the Sun only six diameters of the solar disc ! a 6 THE WORLDS OF SPACE. good proof of the excellence of the instrument as well as the keen sight of the observer. Schiaparelli took the precaution of observing the planet not merely at the same hour on each day, but at various hours on several successive days. On these occasions he could not detect any appreciable change in the position of the spots observed, which apparently remained stationary for long periods at a time. He, therefore, arrives at the remarkable conclusion that the planet rotates on its axis in the same time that it revolves round the Sun namely 88 days, with its axis of rotation nearly at right angles to the plane of the planet's orbit. This remarkable and wholly un- expected result is, however, not without precedent in the solar system. We have a similar case in our moon, which rotates on its own axis in the same time that it revolves round the Earth, in the satellites of Jupiter, in the outer satellite of Saturn, Japetus, and probably also in the satellites of Mars. Assuming the reality of Schiaparelli's discovery, let us see what the result will be of this equality between the period of Mercury's revolution round the Sun, and that of its rotation on its axis. Some may suppose that the length of the day on Mercury will be simply increased in length to 88 times the length of our day, with a regular alternation of day and night. But this is not exactly what will happen. In the case of our Moon, the length of its day (that is, day and night) is certainly equal to its period of ARE THE PLANETS HABITABLE ? *] revolution round the Earth ; but then the Moon revolves round the Earth, not round the Sun ; and although it persistently turns the same face towards the Earth, every portion of its surface will, in the course of a lunar month, be turned towards the Sun. In the case of Mercury, however, as its rotation period is equal to its period of revolution round the Sun, the planet will constantly present the same face to the Sun, as the Moon does to the Earth. The result of this of course will be, that on one side there will be constant day, and on the opposite side per- petual night. At first sight the arrangement would seem to render both sides of the planet uninhabitable. A little further consideration, however, will show that along a narrow zone near the circle dividing the light from the dark side, an inhabitant of Mercury would have the Sun constantly near his horizon, and con- sidering the great intensity of the solar heat on Mercury, compared to that which we experience, this arrangement might serve to render a small zone of Mercury's surface habitable by some forms of life. Owing to the elliptical shape of the planet's orbit round the Sun, its velocity will vary to a considerable extent. We may, however, conclude, from analogy, that the rotational motion on its axis is uniform. This difference in the regularity of the two motions will, of course, give rise to a " libration " similar to the Moon's libration which will have the effect of bringing a portion of the dark side of Mercury 8 THE WORLDS OF SPACE. periodically into the sunlight, and will thus diminish the area of the planet's surface which is shrouded in perpetual night. About three-eighths of the total surface will for ever remain in darkness, three-eighths in perpetual sunshine, while the remaining one-fourth will have alternately day and night. In fact, an inhabitant living near the mean boundary line, and on the planet's equator, would have 44 days of sunshine, and 44 days of night or twilight. A little further in, on the bright side, there would be per- petual day. A little further in, on the dark side, there would be perpetual twilight ; and further in still eternal night would reign. Owing to the low altitude attained by the Sun near the boundary line, its intense heat and light would of course be much mitigated, so that probably this region of the planet's surface may be comparable with the temperate zones of the Earth. Of course in the centre of the bright or dark side we cannot imagine any life could exist ; but at a little distance from the dividing zone, we may suppose the existence of a tropical zone towards the bright side, and of an arctic zone towards the dark side. It may be asked, How did this slow rotation of Mercury probably originate ? It has possibly been caused by tidal friction acting through long ages on the oceans which we may suppose to exist on the surface of the planet. If our Moon was originally covered to a large extent with water, the great mass of the Earth about 80 times the Moon's mass ARE THE PLANETS HABITABLE ? Q combined with the small force of gravity on the Moon (about six times less than gravity on the Earth), would have produced enormous tides, which would, by their friction, have gradually reduced the velocity of its rotation, thereby lengthening the day, and eventually bringing it into equality with the period of its revolution round the Earth. In the case of Mercury, its proximity to the Sun, combined with its small mass, would produce far greater tides than those on the Earth, and the resulting friction would have operated powerfully in reducing the rota- tion velocity of so comparatively small a planet. The evidence in favour of an atmosphere round Mercury seems inconclusive. Of course we cannot imagine the existence of life on a planet utterly devoid of an atmosphere, but if the atmosphere dif- fered very much in constitution from ours, life as we know it would not be possible on the surface of such a planet. From appearances observed on Mer- cury when in the crescent form, Schroter inferred the existence of a dense atmosphere. The correctness of this conclusion has been doubted, but some of Schiaparelli's observations indicate that an atmo- sphere and water probably do exist on the planet. If this be so, the water is probably at or above the boiling-point near the centre of the bright side, and perhaps frozen solid on the dark side. In the case of Venus, we have a planet not differ- ing much from the Earth in size, its mass being about 10 THE WORLDS OF SPACE. four-fifths that of our globe. It is of course much nearer the Sun, and the intensity of the solar light and heat on Venus must be about double what we experience. . The period of rotation on its axis, or the length of its day, is somewhat uncertain. Schroter and De Vico made it about 23 j hours, and observations by Cassini and Madler appear to con- firm this result ; but the exact period would seem to be still an open question. Still greater uncer- tainty is attached to the position of its axis of rotation. Some observers make its inclination to the planet's orbit plane about 73^ ; but this has not been confirmed. Denning finds the markings on Venus much more difficult to see than those on Mercury, usually, indeed, of almost evanescent faint- ness. Under these circumstances the difficulty of determining the period of rotation, and the position of the planet's axis, will be easily understood. That Venus possesses an atmosphere somewhat similar to our own, and of considerable density, seems beyond doubt. The arc of light seen round the disc of Venus before its entrance on the Sun in the recent transits clearly shows refraction of light by a gaseous envelope. Spectroscopic observations made by Pro- fessor Young during the transit of 1882 showed unmistakable indications of an atmosphere surround- ing the planet ; and the observations of Cassini, Madler, Noble, Schroter, Secchi, and others all tend to confirm the evidence afforded by the prism. ARE THE PLANETS HABITABLE? II Indeed Nelson (Neville) considers that it has probably double the density of our own. An attempt has been made to explain the remark- able brilliancy of Venus by supposing that the sunlight is reflected from a persistent cloudy stratum surround- ing the planet. This stratum would of course have the effect of shielding the planet's surface from the intense heat and glare of the Sun ; but it would also suggest an almost constant rainfall a condition rather unfavourable to the existence of animal life. If we suppose the inclination of the axis to the orbit plane to be about the same as that of the Earth, the seasons on Venus will not differ much from ours except in duration, the length of the planet's year being about 225 days, while ours is 365. 1 The shape of Venus's orbit is nearly circular, so that the planet would not, as in the case of Mercury, experience any violent changes of temperature. Owing to the in- creased intensity of the solar heat on Venus we cannot suppose that its equatorial regions can form the abodes of life ; but possibly in the regions sur- rounding the planet's poles the temperature may be sufficiently cool to admit of some forms of animal existence. Indeed, some observations by Gruithuisen and Trouvelot seem to indicate the presence of solar snow-caps, similar to those observed on Mars ; but of 1 Schiaparelli thinks that the rotation of Venus is similar to that of Mercury. If this be so, the above remarks with reference to the planet's seasons will not hold good. 12 THE WORLDS OF SPACE. course observations of this kind are open to consider- able doubt, and the faintness of the markings seen on Venus leaves the question of its physical condition one of much uncertainty. Passing over the planet Earth, which we know to be inhabited by various forms of life even in the heart of " Darkest Africa " we come now to the "red planet" Mars. Of all the planets of the solar system, Mars presents the greatest resemblance to the Earth. The markings on its surface, which have been well observed and mapped, clearly indicate the presence of land and water on its surface. From the observed motion of these spots, the period of the planet's rotation on its axis, or the length of its day, has been accurately determined to a fraction of a single second. This period amounts to 24 hours, 37 minutes, 22f seconds, differing but little therefore from our own day. The inclination of the axis of rotation to the plane of the planet's orbit is, according to Schiaparelli, 65 48', or nearly the same as that of the Earth. The seasons therefore on Mars must be similar to ours, but of longer duration ; the length of the'planet's year being about 687 of our days. Observ- ations seem to show that the atmosphere of Mars is not so dense as ours. That the planet has an atmosphere is, however, beyond all question. This is proved by the presence of snow-caps, by the occasional observation or blurring of the markings by clouds, and by the stronger ARE THE PLANETS IIAfilTAHLE ? 13 evidence of the spectroscope, which shows " air lines " in the spectrum, evidently due to the presence of watery vapour in a gaseous envelope surrounding the planet. Considering, however, the distance of Mars from the Sun greater than that of the Earth in the proportion of 47 to 31 the intensity of the Sun's heat will be much less than the Earth receives. Indeed, it would be reduced to less than one-half of what we experience. Still at the planet's equator, and under a vertical sun at noon, the temperature may perhaps be sufficiently high to support some forms of life. Curiously enough most of the land on the planet's surface seems to be distributed along the planet's equatorial regions, the greater portion of the water surface being collected round the north and south poles. The areas of the land and water on Mars are about equal, whereas with us the water sur- face exceeds that of the land in the proportion of 3 to i. The land and water on Mars are also more evenly distributed than on the Earth, and this would have the effect of equalizing the temperature, and would perhaps make the conditions of life more favourable than we might at first sight be disposed to imagine. Mars must, however, be an undoubtedly cold planet, and the question why its surface is not more covered with ice and snow than it is, is one which has not yet been satisfactorily answered. Possibly some special constitution of its atmosphere may mitigate the severity of its climatic conditions. I am disposed to 14 THE WORLDS OF SPACE. think, however, that if life ever existed on the surface of Mars it has now become extinct. Of the habitability of the minor planets we know almost absolutely nothing. Their small size, however, and their great distance from the Sun would lead us to conclude that the existence of any forms of life on their surface is at least highly improbable. Of the large planets, Jupiter, Saturn, Uranus, and Neptune, observations seem to show that they are still in a highly heated condition, and, therefore, quite unfit for the support of any form of animal life. Neptune, however, appears to be cooler than the others, and possibly its internal heat may be only sufficient to raise the temperature of its surface to a point fitted for the maintenance of life. But this is of course merely conjecture. The heat which Neptune receives from the distant Sun is, however, certainly very small about ^ of tne intensity on the Earth too small, indeed, to have any appreciable effect in rendering the planet habitable. It is almost universally admitted by astronomers and physicists that the Sun is gradually cooling down. That it was hotter in geological times seems clearly indicated by the coal-beds found in the arctic regions, and their existence even in the British Islands is evidence in the same direction. In those far distant times Mars was possibly a habitable and inhabited planet, but has now probably passed the life-bearing stage of its existence, through which our earth is at ARE THE TLANETS HABITABLE ? I 5 present passing. When, in the course of ages, the Sun has still further cooled down, all life will probably cease to exist on our globe, and in that remote epoch Venus will probably form the theatre of life. If life now exists near its poles it will then probably extend to its equator, and the cloudy canopy, in which it now seems to be shrouded, will then, owing to the diminution of the solar heat, be gradually dissolved, and the glories of the starry heavens will be revealed to its wondering inhabitants. Later still, in the march of time, life will die out on Venus also, and then Mercury will become cool enough even at the centre of its sun-lit side to be inhabited by animal life. At last the solar heat being reduced to its minimum, life will cease on Mercury also, and the Sun himself, perhaps, will " roll through space, a cold and dark ball." Such may possibly be the course of life in the solar system. As a writer has well said, " When the birth, the progress, and the history of sidereal systems are considered, we require some other unit of time than even the comprehensive one which astronomy has unfolded to our view. Minute, and almost infinitesimal as is the time which comprises the history of our race, compared with that which records the history of our system, the space even of this latter period forms too limited a standard wherewith to measure the footmarks of eternity," II. TERRESTRIAL AND SUN-LIKE PLANETS. THE planets forming the solar system may be divided into three groups, viz.: I. Th^ Terrestrial Planets, including Mercury, Venus, the Earth, and Mars. 2. The Minor Planets, or asteroids, as they are sometimes called, which form a ring or zone of small planets revolving round the Sun between the orbits of Mars and Jupiter ; and 3. The large planets, sometimes called the Major Planets, which include Jupiter, Saturn, Uranus, and Neptune. The Terres- trial Planets are so called because they present many points of resemblance to the Earth. The length of the day in each except perhaps in Mercury 1 is probably nearly the same. The inclination of the planet's equator to the plane of its orbit is in all very similar. Each probably possesses an atmosphere of analogous composition to our own ; and each ( has 1 Professor Schiaparelli, the eminent Italian astronomer, concludes from his own observations that Mercury rotates on its axis in the same time that it revolves round the sun, viz. about 88 days. TERRESTRIAL AND SUN-LIKE PLANETS. 1 7 probably its surface diversified by land and water. In the case of the Earth we have the equatorial regions, although very warm, fairly habitable by living creatures, while the regions immediately sur- rounding the poles are probably devoid of all animal life. This state of things may possibly be reversed on Mercury and Venus. In these planets, where the total light and heat are so much greater than we receive, the equatorial regions are most probably uninhabitable owing to the intense heat, while the regions round the poles may be sufficiently cool to form the abodes of life. This is very probably the case with Venus in particular, where the total heat received from the Sun is not more than double the amount we have on Earth. The excessive brightness of its surface found by one astronomer to be about ten times as bright as the surface of the full moon ! and reflecting, according to Zollner, as much light as freshly fallen snow would imply that most of its light is reflected from a persistent stratum of dense clouds which would of course reflect considerably more sunlight than the surface of the planet itself. This heavy cloud stratum would mitigate to a con- siderable extent the intensity of the solar rays, and if sufficiently dense and persistent, then to the inhabitants of Venus, if any there be, the Sun and stars will for ever remain unseen by them, for, like a pall, the cloudy canopy covers all. Mercury must, of course, be very much hotter than Venus, but still its 1 8 THE WORLDS OF SPACE. polar regions may perhaps be tolerably cool, and possibly physical conditions, of which we know little, may serve to moderate the intense heat to which its surface is subjected. As, however, Mercury is much less luminous than Venus its atmosphere is probably not so cloud-laden. This is confirmed by the observations of Mr. Denning, who finds that the markings on Mercury are much more easily seen than those on Venus, and Zollner found that it reflects only \2\ per cent, of the solar rays, very similar therefore, in this particular, to the moon, which reflects about 17 per cent, of the incident solar light. Mars we know possesses land and water, an atmosphere, and length of day similar to ours, and seasons which, though of longer duration, are not widely different from those which we experience. Its supply of heat is of course much less, but still its equatorial regions, assisted by special physical con- ditions, may possibly be much warmer than we might at first sight be disposed to imagine. The more equable distribution of land and water on its surface, which we know exists on Mars, would also tend to equalize the temperature, and possibly render the planet at least in regions near the equator a fitting abode for some forms of animal life. Probably, however, a slight excess of heat would be more favourable to animal existence than a deficiency of solar warmth, and on the whole I am inclined to believe that of the planets of the Solar System, Venus. TERRESTRIAL AND SUN-LIKE PLANETS. IQ is the one most likely to be inhabited by sentient beings like ourselves. Next to Mars, we have a zone or ring of small planets, of the physical conditions and habitability of which we know little or nothing. All we know about them is that they are very small, and therefore unlike the terrestrial planets on the one hand, and the " giant " planets on the other, between which groups they are situated. It was suggested by Olbers that they may possibly form the fragments of a larger planet which has been shattered into pieces by the force of some internal explosion. This somewhat plausible hypothesis at one time gained some credence, but there are many objections to the theory, and it is not now generally accepted by astronomers. Professor Vaughan has lately suggested as a more probable origin the collision of two planets of " not very unequal size or mass." This would . become possible when, from a long series of perturbations, the orbits became very eccentric, and the aphelion of one orbit came into conjunction with the perihelion of the other, and near the intersection of their planes. Professor Vaughan points out that the distances of the satellites of Saturn from their primary " show a near conformity to geometrical progression " (with a common ratio of 1*30756). He thinks that apparently four satellites are missing from the Saturnian system, and that possibly these may have become asteroids. 20 THE WORLDS OF SPACE. This seems the converse of the hypothesis which has been advanced, that the satellites of Mars were originally minor planets, which, coming too near Mars, were captured by it. 1 When we come to examine the large planets we find a condition of things totally different from the " terrestrial " planets. They are all very large bodies ; very much larger than the Earth in volume ; very much lighter in density than the Earth ; rotating on their axes much more rapidly, and surrounded by atmospheres much more extensive than our own. Considering the small specific gravity of these gigantic planets, we are compelled to consider that their physical condition must be wholly different from that of the Earth. The density of Jupiter does not differ much from that of the Sun. That of Saturn is about O'66 (water being i). That of Uranus about the same as water, and that of Neptune somewhat less. Considering their enormous absolute mass, in the case of Jupiter 312 times, and in the case of Uranus about I4j times the mass of the Earth, and that this mass must act with great force to compress the materials from the surface towards the centre of the planet, we see at once that their con- dition must be utterly unlike that of the Earth. In the case of the Sun its small density is simply and satisfactorily explained by its intense heat, which is sufficient to maintain its surface in the gaseous state, 1 Comfites Rendus, Nov. 28, 1887. TERRESTRIAL AND SUN-LIKE PLANETS. 21 and, according to Helmholtz and Sir W. Thomson, in the fluid state even to its centre. Unless we assume a somewhat similar state of things in the large planets, we are unable to explain why their density should be so small. Let us see what evidence observation affords in support of this view. The "belts" of Jupiter are familiar to almost every one, at least by name. These are darkish markings on the planet's apparent sur- face, usually parallel to the planet's equator, and may be seen with telescopes of moderate power, though, of course, large instruments are necessary to examine their details satisfactorily. These belts are not permanent features, but are subject to great and rapid changes, both in form and position. Now if we consider the great distance of Jupiter from the Sun more than five times greater than that of the Earth and that the heat derived from the great central luminary is enfeebled in the ratio of the square of the distance, we see how little effect the Sun's heat can possibly have in producing changes in Jupiter's atmosphere. Some other cause must therefore be at work, capable of causing changes of such magnitude as to be visible at the vast distance which separates the Earth from Jupiter. Mr. Ranyard, the well-known astronomer, has attempted to show that spots on Jupiter are more prevalent when spots on the Sun are most numerous, and he considers that possibly both phenomena may be due to the same or similar 22 THE WORLDS OF SPACE. cosmical cause. This view seems to be confirmed by Mr. Browning, who finds that the red colour of the belts coincides with the epoch of Sun spot maxima. An enormous reddish spot of elliptical shape, and measuring some 25,000 miles in length by 7000 in breadth, has been visible for some years past on the southern hemisphere of the planet. It was most clearly visible in the year 1882, which was near an epoch of Sun spot maximum. In 1884 it was very faint and difficult to be seen, but it afterwards partly reappeared, and in 1885 was observed to have its central portion apparently covered with a white cloud, thus giving it the appearance of an elliptical ring. Observations in 1888 showed it still visible, though faint. Perhaps the most remarkable fact connected with it is, that the period of Jupiter's rotation on its axis, computed from observations of this spot, has been slowly increasing since the spot first became visible. Now, from the analogy of the Earth and Mars, we may conclude that the rotation period of Jupiter is uniform, at least so uniform during a period of nine years, that no observations could detect any change. The observations therefore of this red spot would seem to indicate that it has a drifting motion of its own over the surface of the planet, and this, combined with the changes in appearance it has undergone, and its enormous magnitude, would imply the action of forces which have no parallel on Earth. TERRESTRIAL AND SUN-LIKE PLANETS. 23 It must be added, however, that some astronomers including Professor Hough and Mr. Lynn are of opinion that the red spot forms a portion of the actual body of the planet, and that consequently the increase in the rotation period is real, and not merely apparent! As the solar heat is evidently incapable of producing the observed changes in the belts and markings on Jupiter, there seems to be no escape from the conclusion that the forces at work have their origin within the globe of the planet itself. Some curious observations have been made of Jupiter's satellites when transiting the disc of their primary. Satellite III. has often been seen quite as black as the shadow it casts on the face of the planet ! On one occasion the American astronomer Bond saw this satellite as a black spot lying between its own shadow and the shadow of I., and not to be distinguished from either shadow except by its position. The shadow of I. was observed to be very faint at Stonyhurst in November 1880, and Gorton saw it grey on one occasion, which seems to imply that the planet has some intrinsic light of its own. In the case of Saturn we have a somewhat similar condition of things. " Belts " are also visible here, but of course, owing to its greater distance from the earth, any changes which may occur in them although pro- bably on quite as gigantic a scale are not so easily observed as those of Jupiter. The markings on 24 THE WORLDS OF SPACE. Saturn are however undoubtedly variable both in size and position, and like those of Jupiter of enormous dimensions. As the mean distance of Saturn from the Sun is greater than that of Jupiter in the proportion of 886 to 483, or about 1 1 to 6, we have the intensity of the solar heat further diminished in the ratio of the squares of these numbers, or as 121 to 36 ; or in other words, the intensity of the heat derived by Saturn from the Sun is less than one-third of that received by Jupiter. We see therefore how small an effect the Sun's heat can have in producing changes in Saturn's atmosphere. Considering also the very small density of Saturn's globe not much more than half that of Jupiter and about that of walnut wood, the argument in favour of inherent heat in Jupiter is further strengthened in the case of the "ringed planet." The immense distance of Uranus from the Sun prevents us from obtaining much evidence respecting its physical condition, and this remark applies of course with greater force in the case of Neptune. Traces of belts have, however, been recently detected on the disc of Uranus, with some of the monster telescopes of modern days, and observations of these markings seem to indicate that the period of rotation does not differ much from ten hours, as in the case ot Jupiter and Saturn. Mr. Taylor has recently exam- ined the spectrum of Uranus with the aid of Mr. Common's five foot reflector at Ealing. He finds TERRESTRIAL AND SUN-LIKE PLANETS. 25 that " the most striking features are four dark bands in the orange, green, greenish-blue, and blue respect- ively, and bright flutings in the red, orange, and green. No trace of any solar lines, or of any narrow lines, could be seen in the spectrum." The latter have, however, been detected by Dr. Huggins, by means of photography and an exposure of two hours. These were found in the violet end of the spectrum, whereas the bright flutings seen by Mr. Taylor were near the red end. Mr. Taylor satisfied himself that the bright flutings were not due to effect of contrast with dark lines, and he found that the Sodium line D was absent from the spectrum. He concludes that " the presence of the bright flutings in the spectrum of Uranus indicates that we must considerably modify our ideas as to the physical constitution of this planet (and most probably of Neptune also), for there can be very little doubt that it is to a large extent self- luminous." Mr. Taylor's observations were confirmed by Messrs. Bicknell, Crossley, and Fowler, using the same instrument. From these observations, Mr. Ranyard considers "that the natural spectrum of Uranus may only be that of a warm or red-hot body shining with long wave lengths. Therefore at the violet end we might have only the solar light, and at the red end an additional light from the body of the planet." It should be stated, however, that observations at the Lick Observatory seem to show 26 THE WORLDS OF SPACE. that the light of the planet is merely reflected sun- light. The spectra of Jupiter and Saturn also present some peculiarities different from that of ordinary reflected sunlight, but not in so remarkable a degree as that of Uranus. If we consider the members of the Solar System to have been originally formed from the condensation of an enormous heated nebulous mass, as in Laplace's Nebular Hypothesis, or by some other process of evolution, as many astronomers suppose, then it would follow that the large planets would owing to their greater size cool down more slowly, and for this reason we should expect to find that the large planets are still hotter than the Earth. We know from the evidence of mines, deep Artesian wells, volcanoes, and hot springs, that even the Earth, which has long since cooled down on its surface, still retains in its interior some of its primeval heat. In the large planets the original heat probably still exists even on their surfaces. We must, however, conclude that, although hotter than the Earth, they are much cooler than the Sun, which, owing to its vastly greater size, of course parts with its heat still more slowly. If we consider that the Sun's heat is due, as seems highly probable, to the shrinkage or condensation of its mass, we may also conclude that the shrinkage of Jupiter's mass must produce a certain amount of heat, which may be sufficient to keep its surface in at TERRESTRIAL AND SUN-LIKE PLANETS. 27 least a red-hot state. In fact, there would seem to be evidence in favour of the theory that these immense planets are in an intermediate state between the Sun and the " terrestrial " planets. If we look upon them as possessing inherent heat, we may suppose them also to have some intrinsic light of their own in addition to the light reflected from the Sun, unless, indeed, they are merely at a dull red heat which would emit but little light. If, however, they do emit any light, we might naturally expect to find that these larger planets would shine with greater bright- ness in our night skies than would be due to their relative distances from the Sun and Earth. Let us see what evidence observation affords us on this point. From photometric observations Zollner found that the reflective power of Mars, or its " albedo," as it is called by astronomers, is 0*2672, or, in other words, that Mars reflects about 26f per cent, of the solar light falling on it. For Jupiter he found an " albedo " of o - 62, that is, a reflective power of 62 per cent, (that of white paper being 70), or more than double that of Mars, and nearly four times that of the Moon, of which he found the albedo only O'i^6. We must recollect, however, that the reflected light from Mars is absorbed to some extent by the planet's atmosphere ; in fact a double absorption takes place, first in passing from the Sun through the atmosphere to the planet's surface, and then back again to the Earth. Making 28 THE WORLDS OF SPACE. due allowance for this, however, it is evident that the brightness of Jupiter is considerably greater than that of Mars, and if we consider that a large portion of Jupiter's disc is darkened by belts and spots, we must conclude that the brighter parts of its surface must be very bright indeed probably considerably brighter than snow. In the case of Uranus compared with Jupiter, we have the mean distances from the Sun represented by the numbers 5*20 and 19*183 respectively, and their apparent diameters 4" and 46". From these data I find that Jupiter should be 1799 times brighter than Uranus. Now the stellar magnitude of Uranus when in opposition being 5*46, and that of Jupiter 2*52, according to Zollner, we have the observed light of Jupiter equal to 1556 times the light of Uranus. Hence we have the albedo of Uranus (assuming Zollner's value for Jupiter) equal to ilfj- XO'62 I 'i 5 X 0*62 = 07 1, a brightness equal to that of white paper! a result which certainly points to intrinsic light in this distant planet, and obviously tends to confirm the spectroscopic evidence of heat sufficient to dissociate water. If we agree to consider the large planets as being in a highly heated condition, we must of course abandon the idea that they can possibly form the abodes of life, and the description of some writers of the surpassing splendour of Saturn's sky, with its TERRESTRIAL AND SUN-LIKE PLANETS. 29 rings and satellites, as viewed by an inhabitant of Saturn, must be given up as a pleasant dream. The case may, however, be very different with the satellites which circle round them, and these giant planets may very possibly play the part of miniature suns to their attendant family of moons. Indeed Saturn's system seems to form a sort of miniature of the solar system. It has eight satellites, corresponding to the eight larger planets revolving round the Sun, and a system of rings now generally admitted to consist of a swarm of minute satellites similar, even in its divisions, to the zone of minor planets revolving between the orbits of Mars and Jupiter. Seen from Jupiter's nearest satellite, the planet's enormous globe would show a disc of about 19 in diameter, or over 1300 times the area of the full Moon. From the other satellites his disc would vary from 12 to about 4. From Mimas, the nearest satellite of Saturn, the planet would present a disc of no less than 33 in diameter, and with the encircling ring system would afford a considerable amount of light, and form a magnificent spectacle in the mid- night sky of Mimas. Seen from the satellites of Uranus the planet would show a disc varying from r 5i to 5 in diameter. We see, therefore, that the amount of light, and possibly heat, received by the satellites of these systems from their primary may possibly form a 30 THE WORLDS OF SPACE. considerable addition to the scanty supply received from the Sun, and that the existence of some forms at least of life on their surface may be more probable than their great distance from the Sun would at first sight lead us to imagine. III. LIFE IN OTHER WORLDS. THE question is often asked, Are the stars in- habited ? To this we can confidently answer, No. The stars themselves are certainly not habitable by any forms of life with which we are familiar. That the stars are luminous incandescent bodies, similar to the Sun, seems almost self-evident. That they shine by their own inherent light, and not by light reflected from another body, like the planets of the Solar System, is a fact which scarcely needs demonstration. There are no bright objects near them from which they could derive their light, and they are too far from the Sun to obtain any illumination from that source. But if any proofs were necessary, we have the evidence of the spectroscope, which shows un- mistakably that their light emanates from incandes- cent bodies. Many of the stars show spectra very similar to that of the Sun. The light of others, although differing somewhat in quality when analyzed by the prism, indicates clearly that they are at a very high temperature in many cases, indeed, suggesting 32 THE WORLDS OF SPACE. that they are actually hotter than the Sun. It may be objected, however, that in the case of binary or revolving double stars, the smaller component may possibly shine by light reflected fr.om the brighter star. Indeed this has been suggested in the case of Sirius and its faint companion. But I have shown elsewhere 1 that, if the companion of Sirius shone merely by reflected light from its primary, it would be much fainter than it is, and, indeed, would be utterly invisible in our largest telescopes. Further, in some double stars, spectroscopic observations suggest that the component stars have different spectra. This is, of course, conclusive evidence againt the hypothesis of borrowed light ; for were the smaller star to shine by reflected light from the larger, the spectra of both would be identical, as in the case of the Sun and Moon. We may therefore conclude that all the visible stars are suns, and totally unfit for the habita- tion of living creatures. But, may not the stars have planets revolving round them, forming solar systems similar to our own ? As they are evidently suns shining by inherent light, may they not form centres of planetary systems ? In the case of those stars having spectra differing from the solar spectrum, we cannot speak with any con- fidence ; but for those which show spectra similar to that of our Sun, and having therefore, probably, a 1 Journal of the British Astronomical Association, March LIFE IN OTHER WORLDS. 33 similar chemical constitution, the existence of planets revolving round them seems, from analogy, very prob- able. I here refer to single stars, that is stars which have no telescopic close companion ; for the double stars may, perhaps, form systems differently consti- tuted. In any case these binary systems would not be strictly comparable with ours, for the Sun is certainly a single star. Whether systems of planets really revolve round the stars referred to, is a question which, unfortunately, cannot be decided by observation. I have shown else- where * that if a planet equal in size to the " giant planet" Jupiter were revolving round the nearest star a Centauri at the same distance from that star that Jupiter is from the Sun, it would be utterly invisible in our largest telescopes. The invisibility of planets circling round the stars is therefore no proof whatever of their non-existence. Each star of the solar type may possibly be attended by a retinue of planets which may perhaps remain for ever invisible in the largest telescopes which man can construct. We can, therefore, draw, our conclusions only from analogy. If other suns exist resembling our own Sun in chemical constitution, which we know to be a fact, is it not reasonable to suppose that they also form centres of planetary systems similar to the solar system ? 1 The Scenery of the Heavens, p. 165. D 34 THE WORLDS OF SPACE. Consult with reason, reason will reply, Each lucid point which glows in yonder sky, Informs a system in the boundless space, And fills with glory its appointed place ; With beams unborrowed brighten other skies, And worlds to the unknown with heat and light supplies. The suns, which we call stars, were clearly not created for our benefit. They are of very little practical use to the Earth's inhabitants. They give us very little light ; an additional small satellite one con- siderably smaller than the Moon would have been much more useful in this respect than the millions of stars revealed by the telescope. They must, there- fore, have been formed for some other purpose. On Laplace's Nebular Hypothesis, the condensation of an original nebulous mass endowed with a motion of rotation would result not only in the formation of a sun, similar to ours, but also in a system of planets revolving round the central body. If, indeed, the primitive nebula had no rotation or motions of any kind, the result would be a sun without planets or satellites; but the motions with which all the stars seem to be animated lead us to suppose that this would be a case of very rare occurrence. We may therefore conclude, with a high degree of probability, that the stars at least those with spectra of the solar type form centres of planetary systems somewhat similar to our own. This being surmised, let us consider the conditions necessary for the existence of life on these planets, LIFE IN OTHER WORLDS. 35 There are various conditions which must be complied with before we can imagine life, as we know it, to be possible on any planet. Perhaps the most important of these is the question of temperature. We know that in the universe a great range of temperature exists, from the cold of interstellar space estimated at about 460 below the freezing-point of water to the intense heat which rages in the solar photo- sphere. In this long thermal scale life is, at least on the Earth, restricted within rather narrow limits. Below a certain low temperature life cannot exist. This point is, however, far above the temperature of space. On the other hand, above a certain high tem- perature a low one, however, compared with the intense heat of the solar surface life is also impossible, at least for highly organized beings, like man and the larger animals. For minute microscopic organisms the scale may, perhaps, be somewhat extended ; but even in its widest limits, the range of temperature within which life is possible is, so far as we know, certainly a narrow one. For the support of life and vegetation, light is also necessary, for without no flowers would bloom, nor corn grow and ripen to maturity. To obtain this supply of light and heat it is necessary that a life- bearing planet should revolve at a suitable distance from, and in a nearly circular orbit round, a central sun. These conditions, it is hardly necessary to say, are fulfilled in the case of the Earth. Were we much 36 THE WORLDS OF SPACE. nearer to the Sun than we are, we should suffer from excessive heat, and were we much further away, we should probably perish from cold. For this reason ' the existence of life on the other planets of the Solar System seems very doubtful. Mercury is probably too hot, and the other planets are certainly too cold, so far as heat from the Sun is concerned, unless, indeed, their internal heat is sufficient to raise the temperature of their surface to a point sufficient for the maintenance of life. Indeed, there is good reason to suppose that in the planets Jupiter, Saturn, Uranus, and Neptune, this internal heat is still so great that life would be quite impossible on their surface. Venus, inside the Earth's orbit, and Mars outside, are the two planets which seem to approach nearest to the required con- ditions. We know that both these planets possess atmospheres somewhat similar to ours, and, in Mars at least, land and water most probably exist on its surface. Venus is of course much hotter than the Earth, and Mars much colder, but possibly the polar regions of Venus, and the equatorial regions of Mars, may form suitable abodes for some forms, at least, of animal and vegetable life. Let us proceed, however, to consider some other conditions necessary for the existence of life on a planet. A suitable temperature is, of course, indis- pensable, but this is not all. There are other con- ditions which must be complied with. The planet must have a rotation on its axis, so that every portion LIFE IN OTHER WORLDS. 37 shall in turn receive its due share of light and heat. Each point on its surface must have its day and night, the day for work and the night for rest. The axis of rotation must not lie in the plane of the planet's orbit, but must have a suitable inclination, so that each hemisphere may enjoy its seasons, summer and winter, " seed-time and harvest," in due course. Further, the velocity of rotation on its axis must not be too rapid. If the Earth rotated in a period of one and a quarter hours, bodies at the equator would have no weight, and life would be impossible in those regions. The planet must also possess a mass sufficient to retain bodies on its surface by the force of gravity. In the case of very small bodies, such as the moons of Mars, and some of the minor planets between Mars and Jupiter, objects thrown into the air would pass away into space never to return. The planet should also have a mean density greater than that of water, other- wise the seas would possess no stability, and destruc- tive waves would quickly destroy all life on its surface. All these conditions are fulfilled in the case of Mars as well as on the Earth. In the planet Saturn, how- ever, the density is less than that of water, and in Uranus and Neptune only slightly greater. The planet must also possess a suitable atmosphere. This is an all-important condition for the support of animal life at least for the existence of man and the higher orders of animals. This atmosphere must con- sist so far as we know of oxygen and nitrogen 38 THE WORLDS OF SPACE. gases mechanically mixed in proper proportions, and with a small quantity of carbonic acid gas. Were the oxygen in smaller quantity than it exists in the Earth's atmosphere, life could not be supported. On the other hand, were it much in excess of its present amount, a fever would be produced in the blood which would very soon put an end to animal life. The presence of other gases in excessive quantities would also render the air unfit for breathing. We see, there- fore, that a comparatively slight change in the com- position of a planet's atmosphere would so far as our experience goes render the planet uninhabitable by any of the higher forms of life with which we are familiar. For the support of life on a planet, water is also absolutely necessary. Without this useful fluid the world would soon become a desert, and life and vege- tation would speedily vanish from its surface. Geological conditions must also be considered. It is clearly necessary for the welfare of human beings at least that the surface soil and rocks should contain coal, iron, lime, and other minerals, substances almost indispensable for the ordinary wants of civilized existence. That all or any of the conditions considered would be complied with in the case of a planet revolving round a star, it is, of course, impossible to say. But when we find stars showing by their spectra that they contain chemical elements identical with those which LIFE IN OTHER WORLDS. 39 exist in the Sun and the Earth, analogy would lead us to suppose that very possibly a planet resembling our earth may revolve round each of these distant suns. I say a planet, for evidently there would be only one distance from the central luminary a distance de- pending on its size at which the temperature neces- sary for the support of life would exist, as in the case of the Earth, over the whole of the planet's surface. For other planets of the stellar system, life would be, if it existed at all, most probably confined to restricted regions of the planet's surface. There would, there- fore, be in each system one planet and only one, especially suitable for the support of animal life as we know it. This is with reference to light and heat. If the other conditions were not complied with, then life would probably not exist even on this one planet. In the case of a star larger than the Sun, the planet should be placed at a greater distance than the Earth is from the Sun, but in this case the length of the year and the seasons would be longer than ours. The star which more nearly resembles the Sun in the character of the light which it emits is the bright star Capella. Arcturus has a somewhat similar spec- trum. But these are probably suns of enormous size, if any reliance can be placed on the measures of their distance from the Earth. Other bright stars with spectra of the solar type are Pollux, Aldebaran, /3 An- dromedse, a Arietis, a Cassiopeiae, a Cygni, and a Ursse 40 THE WORLDS OF SPACE. Majoris. Another star is r/ Herculis. The magnitude of this star as measured with the photometer, is about 3|. A parallax found by Belopolsky and Wagner places it at a distance of 515,660 times the Sun's distance from the Earth. If the Sun were placed at this distance, I find that it would be reduced to a star of the third magnitude. This result would imply that 77 Herculis is a slightly smaller sun than ours ; and a planet placed a little nearer to the star than the Earth is to the Sun might, perhaps, fulfil the conditions of a life-bearing world. The number of stars visible in our largest telescopes is usually estimated at 100,000,000. Of these we may perhaps assume that 10,000,000 have a spectrum of the solar type, and therefore closely resemble our Sun in their chemical constitution. If we suppose that only one in ten of these is similar in size to the Sun, and has a habitable planet revolving round it, we have a total of 1,000,000 worlds in the visible universe fitted for the support of animal life. We may therefore conclude, with a high degree of probability, that among the " multitudinous " stellar hosts there are probably many stars having life-bearing planets revolving round them. IV. THE RELATIVE BRIGHTNESS OF THE PLANETS. THAT the planets shine with very different degrees of brightness is a fact familiar, perhaps, to most people. The great brilliancy of Venus, when favourably situated as a morning or evening star, is well known, and has frequently given rise to the erroneous idea that a new celestial visitor had appeared in the sky. Jupiter, when in opposition to the Sun and high in the heavens, as it is some years, also forms a brilliant object in our midnight sky, and it is closely rivalled in lustre by the " red planet " Mars when nearest to the Earth. The difficulty of detecting Mercury with the naked eye, owing to its proximity to the Sun, is well known. When seen, however, under favourable conditions, this planet shines with considerable brilliancy, but as it can only be seen at its brightest for a few days in the morning or evening sky a little before sunrise or a little after sunset, and then only for a comparatively few minutes in the twilight, it generally escapes the observation of the casual observer. The "ringed 42 THE WORLDS OF SPACE. planet " Saturn usually appears brighter than an average star of the first magnitude, and may be easily distinguished by its dull yellow colour. The light of this planet is of course considerably increased when the ring system is widely open, the bright rings being very luminous ; but when the rings are nearly invisible the brightness of Saturn is much reduced. Uranus is just visible to the naked eye on a clear night when its exact position with reference to neighbouring stars is known, but Neptune is quite beyond the range of unaided vision. These differences in the relative brightness of the planets are due to four causes : (i) The distance of the planet from the Sun ; (2) the distance of the planet from the Earth ; (3) the size of the planet ; and (4) the reflecting power of its surface, or the " albedo " as it is termed. Of these the first three are easily determined by observation, and a simple method of computing the relative albedos of the different planets forms the subject of the present paper. The method of computation is as follows : The brightness of two planets will vary inversely as the square of their distance from the Sun, and directly as the size of the planet's disc as seen from the Earth, or, making due correction for their crescent and gibbous forms, as the square of their apparent diameters measured in seconds of arc. The results of this calculation will represent the relative brightness the two planets should have if both had the same albedo. THE RELATIVE BRIGHTNESS OF THE PLANETS. 43 If, however, one of them appears brighter than calculation indicates, it implies that its reflecting power or albedo is greater than the albedo of the other. As the relative apparent brightness can be measured with a photometer, we have all the data necessary for calculation of the relative albedos. The albedo is generally represented as a decimal fraction. This fraction denotes the proportion of light reflected compared with the amount received from the Sun ; the albedo of a surface reflecting all the light which falls upon it would be represented by unity. Probably, however, no such surface exists, the albedo of even freshly fallen snow being less than unity. The difference of albedo in the planets is in some cases very striking. In 1878, when Mercury and Venus were in the same field of view of the telescope, Nasmyth found that the surface of Venus was at least twice as bright as that of Mercury, although Mercury is nearer to the Sun. He compared Venus to clear silver, and Mercury to lead or zinc. From photometric observations by Pickering and Zollner, the brightness of Venus is nearly as great as if its surface was covered with snow, and Zollner found that the surface of Mercury is comparable with that of the Moon, which has a small albedo. This difference of surface bright- ness is very remarkable when we consider that Mercury is. much nearer to the Sun than Venus. If we suppose that the surface of Venus is covered with a cloudy 44 THE WORLDS OF SPACE. canopy, as has been suggested, this cloudy covering would perhaps account for the planet's great reflecting power, but the dullness of Mercury's surface is difficult to understand. Owing to the uncertainty which exists as to the relative apparent brightness of Venus and Mercury, as viewed with the naked eye, it is not easy to compute correctly their relative albedos. Olbers found Venus, at its greatest brilliancy, 19 to 23 times as bright as Aldebaran, but Plummer estimated it as nine times brighter than Sirius, which would make it about 56 times brighter than Aldebaran. Mercury is perhaps about equal to Aldebaran when at its greatest brilliancy. I compared the planet and the star in June 1874, in India, and found them about equal. Assuming that when Venus is at her greatest brightness, she is distant from the Sun 66,000,000 of miles, and that in this position she subtends an angle of 40" of arc, and taking the corresponding quanti- ties for Mercury as 28,000,000 and 8i" respectively, I find that Venus should appear about four times brighter than Mercury. Taking Venus as 20 times brighter than Aldebaran, we have the albedo of Venus equal to five times that of Mercury. Zollner found for Mercury an albedo of o f i3. My calculation would therefore make the albedo of Venus equal to 0-13 x 5, or 0*65. Zollner found 0*50. The data used in the above computation arc, however, too un- certain to yield an accurate result. THE RELATIVE BRIGHTNESS OF THE PLANETS. 45 For the planets outside the Earth's orbit, let us take Mars as our standard. For this planet, Zollner found an albedo of 0*2672, or about double that of Mercury. Comparing Mars and Jupiter, we have the mean distances from the Sun represented by the numbers 1*523 and 5*20. Their surfaces are therefore illumin- ated by sunlight in the inverse ratio of the squares of these numbers. That is, the solar illumination on Mars is to the solar illumination on Jupiter as the square of 5*20 to the square of 1*523, or as 27*04 to 2*32. The apparent diameter of Mars at mean oppo- sition may be taken at i8"of arc, and that of Jupiter at 46". Hence the illuminated surface of Jupiter is (||) 2 or 6*53 times that of Mars. The relative brightness of the two planets should therefore be r sx r 178 ; that is, Mars should be 178 times brighter than Jupiter. Now Professor Pickering found the stellar magnitude of Jupiter, when in opposition, to be minus 2*52, or about 2j magnitudes brighter than the zero of the scale of magnitudes, and that of Mars minus 2*25. This makes Jupiter 1-2823 times brighter than Mars. But we have seen that Mars should be 178 times brighter than Jupiter. Hence Jupiter is 178 x 1*2823, or 2*2825 times brighter than it should be had it the same albedo as Mars. The albedo of Jupiter must therefore be 0*2672 x 2*2825 = 0*609. Zollner found an albedo of 0*62, but Bond computed that Jupiter emits more light than it receives 46 THE WORLDS OF SPACE. from the Sun. 1 This would suggest that the planet shines with some inherent light of its own, a conclu- sion which has also been arrived at from other considerations. In the case of Saturn, the existence of the bright rings complicates the observations of the planet's brightness. Pickering's photometric measures make it about equal to a star of the first magnitude when in opposition and the rings invisible. Mars is therefore 3-25 magnitudes, or about 20 times brighter than Saturn. Now the relative distances of Mars and Saturn from the Sun are represented by the numbers 1*523 and 9'539- The squares of these are 2^32 and 90*99, which implies that the intensity of the solar light on Mars is 39-2 times that on Saturn. Taking the apparent diameter of Mars at 18", and that of Saturn at 19", we have the apparent diameter surface of Mars ( jf) 2 or Iff- that of Saturn. Mars should therefore be 39*2 x fff, or 35*17 times brighter than Saturn. But it is only 20 times brighter. Hence the albedo of Saturn must be greater than that of Mars in the ratio of 35*17 to 20, or the albedo of Saturn = H^ 1 x 0-2672 = 0*47. Zollner found 0*498 1. I am inclined to think, however, from my own observ- ations, that Saturn, when in opposition and shorn of his rings, is slightly brighter than a star of the first magnitude. If this be so, the albedo would have a somewhat higher value than that computed above. 1 Chambers' Descriptive Astronomy, 3rd edition, p. 117. THE RELATIVE BRIGHTNESS OF THE PLANETS. 47 Coming now to the planet Uranus, we find the highest albedo of all the planets. Zollner found 0*64, or slightly greater than that of Jupiter, but I find a still higher value. The relative distances of Mars and Uranus from the Sun are 1*523 and 19* 183. The squares of these numbers are 2*32 and 367*99. Hence the intensity of the solar illumination on Mars is 34I-|.2., or 158*6 times that on Uranus. Taking the apparent diameter of Uranus at 4", and that of Mars at 18", as before, we have the area of the disc of Mars (-f) 2 , or 20*25 times that of Uranus. Hence Mars should exceed Uranus in brightness 158*6 x 20*25, or 3211*65 times, if both planets had the same albedo. Now Zollner found the stellar magni- tude of Uranus to be 5*46 ; Pickering finds it 5*56, and my own eye observations make it about 5*4. We may therefore safely assume its brightness at 5*5 magnitude. This gives a difference of 7*75 stellar magnitude between Mars and Uranus, and implies that Mars is 1259 times brighter than Uranus. But we have seen that Mars should be 321 1 times brighter, if the surfaces of the two planets had the same reflecting power. Hence it follows that the albedo of Uranus must be L!^!^ or 2-55 times greater than that of Mars. We have therefore the albedo of Uranus = 0*2672 x 2*55 = 0*68, or nearly equal to that of white paper, which is 0*70. .Let us now consider the planet Neptune, for which 48 THE WORLDS OF SPACE. Zollner found an albedo of 0*46. The relative distances of Mars and Neptune are 1*523 and 30*054. This gives the solar illumination on Mars 389-32 times that on Neptune. Taking their apparent diameters as 18" and 2-9" respectively, we have the result that Mars should be 14996-6 times brighter than Neptune. Now Pickering found the stellar magnitude of Neptune to be 7*96, which makes Mars 1 0*2 1 magnitudes, or 12,023 times brighter than Neptune. Hence we have the albedo of Neptune = -TTnr>nr X 0*2672 =* 0*333, a result in striking contrast to the albedo found above for Uranus. I think there can be no doubt that Uranus has the highest albedo of all the planets of the solar system. Comparing it with Jupiter, I find by the same method of computation that the albedo of Uranus is equal to the albedo of Jupiter multiplied by 1*213. Hence with Zollner's value of Jupiter's albedo, 0*62, we have the albedo of Uranus 0*75, a very high value indeed, exceeding that of white paper, which is 0*70, and pointing strongly to the conclusion that Uranus is in a highly heated con- dition, a conclusion which seems to be partly sup- ported by the evidence of the spectroscope. To further test the high albedo of Uranus, let us compare the relative brightness of Uranus and Neptune. According to Professor Pickering's photo- metric measures, Uranus is 5*56 magnitude, and THE RELATIVE BRIGHTNESS OF THE PLANETS. 49 Neptune 7*96. Uranus is therefore 2*4 magnitudes, or 9 f i2 times brighter than Neptune. The relative distances of the two planets from the Sun being 1 9' 1 83 and 30*054, we have the intensity of the solar light on Uranus 2-4545 times that on Neptune. But the areas of the discs are as 4* to (2*9), 2 or as 16 to 8 - 4 1. Hence the brightness of Uranus should be Ufi x 2 '4545> or 4'67 times that of Neptune. Hence it follows that the albedo of Uranus must be ^Jf, or 1*9528 times that of Neptune. Assuming Zollner's value of 0*46 for the albedo of Neptune, we have the albedo of Uranus = 0*46 x i'952S = 0*898 (!). Even with the low value of Neptune's albedo, which I have found above, viz. 0*333, the albedo of Uranus would be 0*333 X 1*9528 = 0*65, a value which still makes its albedo the highest of all the planets. It is difficult to say what the albedo of the Earth itself may be. Possibly it does not differ much from that of the planet Mars. Professor Young gives it as perhaps 0*20. The Moon's albedo is rather low, 0*1736 according to Zollner. It is, however, greater than that of Mercury, which seems to have the smallest reflecting power of all the planets. With reference to the satellites, those of Mars are so small that we have no data for computing their albedos. Professor Pickering's estimates of their diameter were made on the assumption that their albedo is the same as that of Mars itself. E 50 THE WORLDS OF SPACE. Assuming a diameter of 3400 miles for the third satellite of Jupiter, the largest and brightest of the system, and the mean diameter of Jupiter itself at 87,000 miles, we have the area of Jupiter's disc 655 times that of the satellite. If both have the same albedo, Jupiter should therefore be 655 times brighter than the satellite. Now Pickering finds the stellar magnitude of this satellite to be 5*24. This makes Jupiter 776 magnitudes, or 1271 times brighter than the satellite. Hence the albedo of Jupiter must be nearly twice that of the third satel- lite. This result agrees with the fact that dark spots have been observed on the third satellite by several astronomers. The diameter of Saturn's largest satellite Titan is somewhat doubtful, but assuming it at 3000 miles, and its stellar magnitude to be 9-43, as measured by Pickering, the diameter of Saturn being 72,000 miles, I find that the albedo of Saturn would be 2'2 times that of Titan. This would make the albedo of Titan about 0*21, but owing to the uncertainty which exists as to its diameter, this result must be considered as somewhat doubtful. The satellites of Uranus and Neptune are so faint that no satisfactory results could be computed. For the satellite of Neptune, Pickering finds a stellar magnitude of 13*82, or 5-93 magnitudes fainter than its primary. If we take the diameter of Neptune THE RELATIVE BRIGHTNESS OF THE PLANETS. 51 at 36,000 miles, and assume that its albedo is twice that of its satellite, I find that the diameter of the satellite would be about 3300 miles. Assuming the same albedo as Neptune, the diameter would be about 2340 miles. V A DOUBLE PLANET. DOUBLE stars are numerous in the heavens, and double nebulae are not uncommon. Even double comets have been recorded, as in the case of Biela's comet, and the faint companions which have been observed in close attendance upon some of the large comets of recent years. The duplicity of one of the satellites of Jupiter has even been "suspected," but, as far as I know, the suspicion has not been con- firmed. Although many of the planets of the Solar System are attended by satellites, they are usually considered as single bodies. We may, however, perhaps, make an exception to this rule in the case of the Earth and Moon, which have been termed " a double planet " for the following reasons : The Moon's volume compared with that of its primary is greater than that of any other satellite of the Solar System. The volume is about T V of the Earth's volume, and its mass about ^ T of that of the Earth. The volumes of the satellites of the other planets bear a much smaller ratio to the volume of A DOUBLE PLANET. 53 the primary. We need not consider the satellites of Mars, which are very minute bodies, and quite insig- nificant in size compared with their primary. The largest of the satellites of Jupiter has a volume only Tg-Jrnr of that of the " giant planet." The largest of Saturn's satellites, Titan, has probably not more than Toiiu-o f tne volume of Saturn. The exact size of the satellites of Uranus is unknown, but judging from their faintness, we may conclude that their volume is small compared with that of their primary. Even the satellite of Neptune, supposed to be the largest satellite of the Solar System, is probably small compared with the planet. Taking its diameter at 3000 miles, and that of Neptune at 36,000 miles, the volume of the satellite will be only TT V- of Neptune's volume. We see, therefore, that the Moon is comparatively a very large satellite. It is, of course, absolutely smaller than the largest satellite of Jupiter, Saturn's satellite, Titan, or the satellite of Neptune ; but compared with the Earth, which is a small planet (in comparison with Jupiter, Saturn, Uranus, or Neptune), it must be considered as really an enor- mous satellite, and in relative size deserving to rank rather as a small planet accompanying the Earth in its annual journey round the Sun, than as a satellite revolving round it. Seen from Venus, the Earth and Moon will appear more like a " double planet " than a planet with an 54 THE WORLDS OF SPACE. attendant satellite. From a consideration of the brightness of Venus as seen from the Earth, we may form an estimate of the probable brightness of the Earth and Moon as viewed from Venus. To do this it will, of course, be necessary to make some as- sumptions. We 'should require, in the first place, to know the albedo, or reflecting power, of the Earth's surface. To determine this accurately would not be an easy matter, but if we assume that it has the same albedo as the Moon, we may not, perhaps, be very far from the truth. Now Zollner found the albedo of Venus to be represented by the frac- tion 0*50, or about three times the Moon's albedo Co- 1 736). Venus, when at her greatest brilliancy, and approach- ing inferior conjunction, is distant from the Earth about 39,000,000 miles, and has then about one- fourth of the area of her disc illuminated by sunlight. The Earth when in "opposition," and therefore at its brightest as seen from Venus, is distant from the planet about 26,000,000 miles. Hence we have the relative distances in the ratio of 39 to 26, or as 3 to 2. If, to simplify the calculation, we assume the diameter of the Earth and Venus as equal, the apparent areas of their discs will be as 3 2 to 2 2 , or as 9 to 4. That is, the area of the Earth's disc when in " opposition," as seen from Venus, will be about 2\ A DOUBLE PLANET. 55 times the area of Venus's disc when at her brightest as seen from the Earth. Now as the Earth shows a full face to Venus when at its brightest, and Venus only one-fourth of a fully illuminated disc when at its brightest to us, we should have the Earth brighter than Venus in the proportion of 36 to 4, or as 9 to i, if the distances of both planets from the Sun and their albedos were the same. But as their dis- tances from the Sun are in the ratio of 93 to 67, Venus will be more brilliantly illuminated in the ratio of 93 2 to 6; 2 , or about as 19 to 10, and as its albedo, as assumed above, is three times greater, we have the brightness of Venus's surface greater than that of the Earth's surface in the ratio of 57 to 10. Hence, finally, we have the brightness of the Earth, when in " opposition," as seen from Venus, brighter than Venus at its greatest brilliancy as seen from the Earth, in the ratio of 90 to 57. Taking the diameters of the Earth and Moon as 7912 miles and 2163 miles respectively, the areas of their apparent discs would be in the ratio of 13-38 to i. Hence, with the same albedo, the Earth and Moon, as seen from Venus, would differ in brightness by 2'S i stellar magnitudes. Now Plummer found that Venus at its greatest brilliancy is nine times brighter than Sirius. The Earth, therefore, as seen from Venus, would appear (24f^) 14-21 times, or 2'88 stellar magnitudes brighter than Sirius. The Earth and Moon would therefore 56 THE WORLDS OF SPACE. shine as two stars, one about half as bright again as Venus at her brightest, and the other about equal to Sirius, and separated, when the Moon is in " quad- rature," by about 31' of arc, forming a superb "naked eye double star," perhaps the finest sight in the planetary system. They would present the appear- ance of a " double planet," in striking contrast with the faintness of the other satellites of the Solar System. The Earth would show a disc of about 62" in diameter, and the Moon one of about 17", and the markings on both might be well seen with a good telescope. Seen from Mars, the Moon would also be visible as a small attendant planet to the Earth, but varying considerably in brilliancy owing to its phases. The Moon's title to rank as a planet rather than a satellite is strengthened by the fact that her path in space is, like the planetary orbits, always concave to the Sun. Professor Young says, in his General Astronomy, that " if we represent the orbit of the Earth by a circle of 100 inches radius, the Moon would only move out and in a quarter of an inch, crossing the circumference 25 times in going once round it." This is a very different arrangement from the satellites of Jupiter and Saturn, which seem to form miniatures of the Solar System. VI. ALPHA CENTAURI AND THE DISTANCES OF THE STARS. THE saying of Seneca with reference to the impossibility of achieving immortality by ordinary efforts, that there is no easy way from the earth to the stars Nan cst ad astra mollis a terris via is one which may be applied in a literal sense to the determination of stellar distances. In old times, Hook, Flamsteed, and Cassini made numerous but unavailing efforts to measure the distance of some of the stars, and it is only in recent years that careful measurements made with accurate instruments have partially solved the enigma. It was during a series of observations made by Sir William Hcrschel, at the close of the eighteenth century, carried out with a view to finding the dis- tance of certain double stars, that he made his great discovery of binary or revolving suns. Although unsuccessful in his efforts to find the distance of the stars in question, his labours were fully rewarded by the discovery of stellar systems moving in obedience 58 THE WORLDS OF SPACE. to the laws of universal gravitation. This important discovery one of the most interesting of modern times seems to have diverted his attention from his original design, but in any case his instruments were not sufficiently accurate for so delicate an investigation. The bright southern star a Centauri is so far as we know at present certainly the nearest fixed star to the Earth. As might be expected from its com- parative proximity to our system, it is one of the brightest stars in the sky. It ranks third in the order of brightness, Sirius being facile princeps, Canopus second, and a Centauri third. It is slightly brighter than Arcturus which may perhaps be con- sidered the leader of the Northern Hemisphere. The idea that this bright star might possibly lie within measurable distance was suggested by two facts. First, by its being a remarkable binary star, with the distance between the components unusually large for an object of this class ; and secondly, from its large " proper motion " across the face of the sky, a fact which is usually assumed to indicate nearness to our system. An attempt to find its distance was made by Professor Henderson in the years 1832-33. Using a mural circle with a telescope of four inches aperture and a transit of five inches, he obtained an "absolute" parallax of 1-14" of arc with a probable error of one-tenth of a second, indicating a distance from the Earth of about 181,000 times the THE DISTANCES OE THE STARS. 59 Sun's distance from the Earth. It may here be explained that the " parallax " of a fixed star is a change in the place of the star, due to the Earth's orbital revolution round the Sun. It is one-half the total displacement of the star as seen from opposite points of the Earth's orbit. In other words, it is the angle subtended at the star by the Sun's mean distance from the Earth, or the radius of the Earth's orbit. The " absolute " parallax is the actual parallax of the star. A " relative " parallax is the parallax with reference to a faint star situated near the brighter star, and which is assumed to lie at a much greater distance from the Earth. Sometimes the "relative parallax " comes out a negative quantity. This implies that the fainter or "comparison star" is nearer to the Earth than the brighter star. Further measures of a Centauri made by Hender- son and Maclear, in the years 1839 and 1840, with two mural circles of four inches and five inches, yielded an absolute parallax of 0-913 of a second, indicating a distance of about 226,000 times the Sun's distance from the Earth, or about 21 billions of miles. A re-discussion of these measures afterwards gave a parallax of 0*976 of a second. From observations in 1860-64, Moesta found with a transit circle of six inches aperture, a parallax of O'88 of a second. From a new determination, the same astronomer found a smaller parallax of 0*521 of a second. Elkin and Maclear, in 1880, re-discussing Maclear's ob- 60 TITE WORLDS OF SPACE. servations, found a parallax of 0*512 of a second. Dr. Gill, in iSSi 1882, using a heliometer of 4] inches aperture, obtained a relative parallax of 076 of a second, with a probable error of only 0*013". Elkin, using the same instrument in the years 1881 1883, obtained a relative parallax of 0*676 of a second. The difficulties attending the measures of an absolute parallax are so great that the relative parallaxes found for a Centauri are now considered the most trustworthy. Assuming that the small comparison stars used in determining the " relative " parallax are at such a distance that their parallax is inappreciable as is probably the case we may assume that the relative parallax is practically the same as the absolute parallax. Dr. Gill's result of 076" for the parallax of a Centauri is now generally accepted as the most reliable. This places the star at a distance of 271,400 times the Sun's distance from the Earth, or about 25 billions of miles, a distance which light, with its great velocity of 186,300 miles per second, would take 4*287 years to traverse. Taking the proper motion of a Centauri at 37" of arc per annum, a parallax of 076" would denote an annual motion of 4*868 times the Sun's distance from the Earth, or a velocity of about 14^ miles per second in a direction at right angles to the line of sight. As there may also be and probably is a motion in the line of sight, either towards or away THE DISTANCES OF THE STARS. 6 1 from the Earth, the star's actual velocity through space is probably greater. As has been already mentioned, a Centauri is a remarkable binary or revolving double star. Its duplicity seems to have been first noticed by Richaud in 1690. Since the year 1752, numerous measures of the position of the components and the distance between them have been recorded, and many attempts have been made to compute the orbit. The apparent ellipse is a very elongated one, and the distance has varied from about 22" to I J". At present the distance is about 21", so that the components may be seen with any small telescope. Unfortunately the star is not visible in these latitudes, but it must form a splen- did telescopic object in the Southern Hemisphere. Various periods of revolution have been assigned to this magnificent pair of suns, ranging from 75?; to 88J years. A recent investigation 1 by Dr. T. J. J. See of Chicago appears to definitely fix the period at 81-07 years. His results agree closely with those found by Dr. Gill, 2 and also with an orbit computed by Mr. A. W. Roberts, 3 and cannot be far from the truth. Assum- ing a parallax of 075", Dr. See finds that the com- bined mass of the components is 1-998 times, or sensibly twice the mass of the Sun. He also finds 1 Monthly Notices, Royal Astronomical Society, December 1893. - Ibid.) Royal Astronomical Society, vol. xlviii. p. 15, 3 Astronomische Nachrichten^ No. 3175, 62 THE WORLDS OF SPACE. that the longer axis of the real elliptical orbit is 23*592 times the Sun's distance from the Earth, or " about a mean between those of the planets Uranus and Neptune ; but the orbit is so eccentric that in Periastron the two stars are only slightly remoter than the Sun and Saturn (11*3), while in Apoastrou the distance considerably surpasses that of Nepttme from the Sun (36-0)." According to Dr. Gould there is a difference of 2^,- magnitudes in brightness between the component stars of a Centauri. This makes the primary star ten times brighter than the companion. If we assume that both bodies have the same density and intrinsic brilliancy of surface, this ratio of brightness would imply that the mass of the larger star is about 31 -J times the mass of the smaller. The spectrum of a Centauri is, according to Pro- fessor Pickering, of the second or solar type, so we may perhaps conclude that it is a somewhat similar Sun to ours, the primary star having a mass nearly twice the mass of the Sun, and consequently a some- what larger diameter. Next in order of distance to a Centauri comes a small star numbered 21,185 in Lalande's catalogue, for which Winnecke found a parallax of about half a second of arc. The distance of this star is, however, not so certain as that of the famous star 61 Cygni, which is generally supposed to be the nearest star in the Northern Hemisphere. Although a comparatively THE DISTANCES OE THE STARS. 63 insignificant star, of about the fifth magnitude, the attention of astronomers was attracted to it by its large " proper motion," about 5*2 seconds of arc per annum, a motion which places it fourth in the order of swiftly moving stars. Numerous measures of its distance have been made by various astronomers, from Arago and Mathieu in 1812, down to Professor Pritchard in 1886 1887. Most of these measures give a parallax ranging from about 0*27 to 0*56 of a second of arc. The mean of recent measures which are of course the most reliable may be taken at 0*45", indicating a journey for light of about 7~| years. This parallax combined with the star's proper motion of 5*2" indicates a velocity of 34 miles a second, at right angles to the line of sight. Like a Centauri, 61 Cygni is a, wide double star, both components apparently moving together through space. This fact evidently points to a physical con- nection between the two stars, and suggests that one revolves round the other, or rather both round their common centre of gravity. Several attempts have been made to determine an orbit, but as the angular motion since their discovery has not been considerable, there is still a doubt as to the binary character of the pair. If they are really revolving, the period of revolution must be measured by hundreds of years. Assuming a period of 782! years found by Peters, I find that the combined mass of the components would be 0-461 of the Sun's mass, with a mean distance 64 THE WORLDS OF SPACE. between them of 65*62 times the Sun's mean distance from the Earth. This result may not be very far fron the truth, for I find that the Sun placed at the dis tance of 61 Cygni would shine as a star of about 2*8 magnitude. 1 Now, taking the magnitude of 61 Cygni at 4*98 as measured with the wedge photometer at Oxford we have a difference of 2*1 8 magnitudes, which implies that the Sun is about 7^ times brighter than the combined light of the components of 61 Cygni, and its mass, therefore, probably considerably greater. Next in order of distance to 61 Cygni comes the brilliant Sirius. Details respecting the distance and probable size of this star will be found in the article on ' Sirius and its System ' in the present volume. For the third magnitude star 77 Herculis, Belopolsky and Wagner found a parallax of 0*40 of a second, or about the same as that of Sirius, but so far as I know this result has not been confirmed by any other astronomer. For the binary star rj Cassiopeia, Schweizer and Socoloff found a parallax of 0*3743 of a second. With this parallax, and assuming a period of 222 years found by Dr. Doberck, I find the mass of the system only 0*366 of the Sun's mass. Placed at the distance of 11 Cassiopeiae, the Sun would, I find, be reduced to a star of 3*2 magnitude, or slightly brighter than the 1 See chapter on ' The Sun among his Peers. 3 THE DISTANCES OF THE STARS. 65 star appears to us. As the spectrum of ?? Cassiopeiae is of the second or solar type, the two bodies may, perhaps, be comparable in physical constitution, and a comparison of their relative brightness agrees fairly well with their relative mass. There are some other stars with fairly well determined parallaxes of ~ to J of a second of arc, but those referred to above are the most remarkable. That a Centauri and the other stars we have been considering are comparatively near neighbours of our system may be seen from the fact that Dr. Elkin finds an average parallax of only 0*089 of a second for stars of the first magnitude. This gives an average distance of 8f times the distance of a Centauri, and implies that an average star of the first magnitude is about 72 times brighter than a first magnitude star placed at the distance of a Centauri. Our nearest neighbour is, however, about twice as bright as an average star of the first magnitude. It follows, there- fore, that, on the average, stars of the first magnitude are really some 36 times brighter than a Centauri. If of the same intrinsic brilliancy of surface, this result would indicate that stars of the first magni- tude are suns about six times the diameter of a Centauri, and, therefore, much larger in volume than that star and our Sun. The theory that the stars may be assumed to be, generally speaking, of nearly equal size and bright- ness, an hypothesis advocated by Sir William Herschel 66 THE WORLDS OF SPACE. and the elder Struve, is now shown to be erroneous by the fact that comparatively faint stars like 61 Cygni and Lalande 21,185 are at a measurable distance from the Earth, while the bright southern star Canopus second only to Sirius in brilliancy is at such a distance that a small parallax of only 0*03 of a second, found by Dr. Elkin, seems of very doubtful value. If the result found by Dr. Elkin for the average parallax of stars of the first magnitude is correct, we are led to the conclusion that the brightest stars in the heavens with the exception of Sirius and a Centauri, and perhaps Procyon owe their brightness to enormous size, and not to comparative proximity to our system. The distances of two stars from the Earth being known, it is easy to calculate the distance between them in space. For, knowing the exact position of each star on the celestial vault, we can compute the angular distance between them. We have then two sides of a triangle and the included angle, and we can, therefore, calculate by trigonometry, or by a simple graphical construction, the length of the third side, which is the required distance between the stars. Taking Sirius and a Centauri, I find that the angular distance between them is 88^ degrees. Now, taking the parallax of a Centauri at 076", and that of Sirius at 0*39", I find that the distance between the two stars is about 589,000 times the Sun's distance from the Earth, This corresponds to a parallax of 0-35 THE DISTANCES OF THE STARS. 67 It follows, therefore that Sirius seen from a Centauri would appear nearly as bright as we see it ; while a Centauri viewed from Sirius would be diminished in brilliancy, and probably reduced to nearly a star of the second magnitude. VII. THE SUN AMONG HIS PEERS. THE Sun is a star, and the stars are suns. This fact has been a familiar one to astronomers for many years, and is probably known to most of my readers. That the stars shine by their own inherent light, and not by light reflected from another body, like the planets of the Solar System, may be easily proved. That many of them at least are very similar to our own Sun is clearly shown by several considerations. I will mention three facts which prove this conclusively. First, their great intrinsic brilliancy compared with their small apparent diameter, a diameter so small that the highest powers of the largest telescopes fail to show them as anything but mere points of light without measurable magnitude. Second, their vast distance from the Earth, a distance so great that the diameter of the Earth's orbit dwindles almost to a point in comparison. This accounts satisfactorily for the first fact. Third, the spectroscope that unerring instrument of modern scientific research shows that the light emitted by many of them is very similar to THE SUN AMONG HIS PEERS. 69 that radiated by the Sun. Their chemical and physical constitution is, therefore, probably analogous to that of our central luminary. The red stars certainly show spectra differing considerably from the solar spectrum, but these objects are comparatively rare, and may perhaps be considered as forming exceptions to the general rule. The stellar spectra have been divided into four types or classes. The first class includes stars like Sirius, in which the strong development of the hydrogen lines seems to indicate the preponderance of this gaseous metal in the glowing envelopes of these distant suns. The second class includes stars in which the spectrum closely resembles the solar spectrum. The third and fourth types include those which show a banded spec- trum, the rainbow-tinted streak of ordinary stellar spectra being crossed by a number of dark bands or shadings, in striking contrast to the solar spectrum, in which fine lines only are visible. These are mostly of an orange or red colour of various degrees of inten- sity, and many of them are variable in their light. There is some reason to suppose that stars of the first type are probably the hottest and intrinsically the brightest of all, and are not, therefore, fairly compar- able with our Sun. In considering, therefore, the Sun's rank in size and brightness among the stellar hosts, we should compare it with those which show a similar spectrum. But how are we to compare the Sun with any star ? 70 THE WORLDS OF SPACE. It is clear that the first thing we require to know is the star's distance from the Earth. The apparent size and brightness of an object depends on its distance from the eye. A candle placed a few feet from us will look larger and give more light than a brilliant electric lamp several miles away. Venus is, at its brightest, considerably brighter than Jupiter, although the former is a much smaller planet than the latter. Unfortu- nately the distance of but few of the fixed stars from the Earth has been ascertained with any approach to accuracy. Failure in the attempt to measure the distance of a star implies, of course, that it lies at a vast distance from the Earth. In several cases, how- ever, the efforts of astronomers have been rewarded with success, although the result found for some stars is still open to much uncertainty. In addition to their distance, we also require to know the apparent bright- ness of the Sun with reference to the star with which it is to be compared. Owing to the excessive brilli- ancy of the Sun compared with even the brightest stars, this is a matter of no small difficulty. Photo- metric measures, made with the aid of the Moon as a " medium," have however yielded a fairly reliable result, and it is now generally assumed by astronomers that on the scale of stellar magnitudes which represents the brightest stars as of the first magnitude, and those near the limit of ordinary eyesight as sixth magnitude, the Sun's light may be expressed as about 26J magni- tudes brighter than an average star of the first magni- I THE SUN AMONG HIS PEEKS. 7 1 tude, such as Altair or Spica. This may seem to some a rather surprising result. It may be asked, if there is a difference of five magnitudes between a sixth magnitude star and one of the first magnitude, should not the difference between a first magnitude star and the Sun be much more than 26J magnitudes ? At first sight, the number representing the Sun's stellar magnitude certainly does seem small, but a little con- sideration will soon dispel this feeling of surprise. The explanation of the apparent difficulty is a simple one and will be easily understood by those familiar with the rules of arithmetic. The numbers denoting star magnitudes really form a geometrical series. Thus a star of the fifth magnitude is about two and a half (more correctly 2*512) times brighter than a star of the sixth magnitude ; a star of the fourth two and a half times brighter than one of the fifth, and so on. This series increases very rapidly, like the question of the nails in a horse's shoes in books on arithmetic. With the " ratio " of 2-512, a star of the first magnitude would be one hundred times brighter than one of the sixth. A difference of ten magnitudes between two stars would denote that one is 10,000 times brighter than the other ; and if we go on to 26J magnitudes above the first magnitude we arrive at a very large number indeed. In fact, the number 26^ implies that the Sun is equal in brightness to 39,81 1,000,000, or nearly 40,000,000,000 of stars of the first magnitude, like Altair or Spica. 72 THE WORLDS OF SPACE. Knowing, then, the Sun's stellar magnitude, we can easily calculate what its apparent brightness would be if placed at the distance of a star of which the distance from the Earth has been determined. For, as light varies inversely as the square of the distance, we have simply to express the distance of the star in terms of the Sun's distance from the Earth, square this number, and then find how many stellar magni- tudes would give the diminution of light indicated by the number thus obtained. A " parallax " of one second of arc would represent a stellar distance of 206,265 times the Sun's mean distance from the Earth. At this distance the Sun would shine as an average star of the first magnitude. If the star's parallax is only a fraction of a second as it always is we have to divide 206,265 by the parallax to obtain the distance sought. For example, the most reliable measures give a parallax for Sirius of about four-tenths of a second of arc. Dividing this into 206,265, we have the distance of Sirius equal to 515,662 times the Sun's distance from the Earth. I find that the square of this number represents a diminution of light of 28J stellar magnitudes. Subtracting 26^ from this, .we have the result that the Sun's light would be reduced to two magnitudes below the first, or to the third mag- nitude, if it were placed at the distance of Sirius. In other words, Sirius, which is about two magnitudes brighter than an average first magnitude star, is four stellar magnitudes, or about forty times brighter than THE SUN AMONG HIS PEERS. 73 the Sun would be in the same position as seen from the Earth. From observations of a faint companion which revolves round Sirius in a period of about 58 years, I find that the combined mass of this brilliant star the brightest of the stellar hosts and its companion, is about three times the mass of the Sun. Now, if Sirius were of the same intrinsic brightness as the Sun, and of the same density, its diameter would be 6*32 (the square root of 40) times the Sun's diameter, and its mass would be 6^32 cubed, or 253 times the mass of the Sun. We see, then, that Sirius is enormously bright in proportion to its mass, or in other words, that it is a much less massive star than its great brilliancy would lead us to imagine. It must therefore differ considerably in its physical constitution from that of our Sun. Other stars of the same class are probably comparable with Sirius in the exceptional brilliancy of their luminous surface. Stars of the first type are, therefore, of probably small mass in proportion to their brightness, and cannot be fairly compared with the Sun in size, or at least in the quantity of matter they contain. Professor Pickering finds that the brightest stars of the Milky Way belong to the Sirian type, and Dr. Gill concludes, from an examination of Galactic photographs, that the smaller stars composing the Milky Way are for the most part blue stars, and have probably spectra of the Sirian type. If this be so, they are prob- 74 THE WORLDS OF SPACE. ably really as well as apparently small, a conclusion which had been previously arrived at from other considerations. Let us now consider stars of the second or solar type. Among the brighter stars of this class we have Capella, Arcturus, Aldebaran, Pollux, a Cygni, a Arietis, a Cassiopeiae, etc., in the Northern Hemi- sphere, and Canopus and a Centauri in the Southern. For Capella, which rivals in brightness Arcturus and Vega (and forms with them the most brilliant trio in the Northern Hemisphere), Dr. Elkin finds a parallax of only slightly more than one-tenth of a second of arc. At the distance indicated by this result nearly 2,000,000 times the Sun's distance from the Earth the Sun would shine as a star of only the sixth magnitude. This implies that Capella is about 250 times brighter than the Sun. If of the same intrinsic brilliancy of surface, its diameter would, therefore, be about sixteen times the Sun's diameter, or nearly fourteen millions of miles ! As the spec- trum of Capella is almost identical with the solar spectrum, it seems probable that the physical con- stitution of the Sun and star are similar. We must, therefore, if its measured distance be reliable, con- sider Capella to be a vastly larger body than our Sun. The above diameter would imply a volume equal to 4000 suns, a truly stupendous globe ! A minute parallax of about one-sixtieth of a second of arc, found for Arcturus by Dr. Elkin, gives a still 1 THE SUN AMONG HIS PEERS. 75 more astounding result. This small parallax implies a distance from the Earth equal to about twelve million times the Sun's distance. This vast distance would produce a diminution of light of about 35 1 magnitudes, so that the Sun placed at the distance of Arcturus would be reduced to a star of only 9! magnitude ! It would not be visible with an opera glass ! Arcturus is, therefore, in round numbers 9^ magnitudes, or over 6000 times brighter than the Sun would be at the same distance. Assuming the same density and brightness of surface as the Sun, the diameter of Arcturus would, therefore, be about 79 times the Sun's diameter, or over 68,000,000 miles, and its mass about 500,000 times the mass of the Sun ; figures well calculated to " stagger the imagination." From the small value of the parallax found for Arcturus, we cannot, of course, place very much reliance on its accuracy ; but there can be little doubt that the distance of this bright star is really very great, and that consequently it is a much larger sun than ours, probably one of the most massive bodies in the universe. A mean of the results found by Elkin and Hall for Aldebaran would reduce the Sun to a star of nearly the sixth magnitude at the same distance, and its light would fade to a star of below the eighth magnitude if it were removed to the distance found by Professor Pritchard for a Cassiopeiae. For the bright star Pollux, Dr. Elkin found a 76 TIFK WORLDS OK STACK. parallax of only O'o68 of a second, representing a distance at which the Sun would be reduced to a star of about the seventh magnitude. This makes Pollux 164 times brighter than the Sun, indicating a diameter about thirteen times greater, or about 11,000,000 miles ! Dr. Elkin's result for the bright southern star Canopus would give the Sun a magnitude of only SJ, if placed at the same distance. As this brilliant star second only to Sirius in lustre is nearly one magnitude brighter than Arcturus, we see that it is probably comparable with the Northern star in size. A negative parallax found by Elkin, Glasenapp, and Peters for a Cygni, and a similar result arrived at by Downing and Main for y Draconis, indicates, of course, that these stars lie at a vast distance from the Earth, a distance, perhaps, too great for our present methods of measurement. Their comparative brilliancy, especially that of a Cygni, would, therefore, suggest that they are very massive bodies, far exceeding our Sun in absolute size. The results I have given will show that the brilliancy of some at least of the brighter stars may probably be explained by their enormous size in comparison with the Sun. Placed at the same distance from the Earth, the Sun would dwindle to an insignificant star, invisible in some cases to the naked eye ! For some stars of the solar class, however, smaller distances have been found. For ?/ Herculis, a star THE SUN AMONG HIS PEERS. 77 of about 3 J magnitude, Belopolsky and Wagner found a parallax of four-tenths of a second, indicating a distance about the same as that of Sirius. As at this distance the Sun would be only reduced to the third magnitude, it would seem that we have here a star of rather smaller mass than our Sun. In the case of binary or revolving double stars, if we can determine their distance we can easily calculate the combined mass of the components in terms of the Sun's mass. Assuming the most reliable distance and the best orbits computed for the following binary stars rj Cassiopeiae, 40 Eridani, Sirius, Castor, a Ccntauri, 70 Ophiuchi and 61 Cygni I find the total mass of these seven stellar systems equal to 14 \ times the mass of the Sun, or an average of twice the Sun's mass for each system. Omitting Sirius and Castor, which have spectra of the first type, the others being of the second, we have a total mass of five systems of ill times the mass of the Sun, or an average of 2-31 for each system. Here we have five suns or rather five pairs of suns, not differing greatly from our own Sun in mass. Indeed one of them, 61 Cygni, is of smaller mass, if the orbit computed by Peters can be relied upon. There seems, however, to be still some doubt as to whether this famous pair really forms a binary system. Its distance from the Earth has, however, been satisfactorily determined by several astronomers. The later results are fairly accordant, and it may be confidently assumed that its parallax 78 THE WORLDS OF SPACE. is about 0*45 of a second of arc, representing a distance of 458,366 times the Sun's distance from the Earth. At this distance I find that the Sun would be reduced to a star of about 2'8 magnitude. Now from the photometric measures made at Oxford, the stellar magnitude of 61 Cygni is 4*98. The difference, or 2' 1 8 magnitudes, implies that the Sun is about 7! times brighter than the combined light of the com- ponents of 6 1 Cygni, and its mass, therefore, probably greater. At the distance of a Centauri the nearest of all the fixed stars the Sun would be reduced to 17 magnitude, or about one magnitude fainter than the star appears to us. This would indicate that, if of the same brightness and density, the mass of the system of a Centauri is about four times the mass of the Sun. A calculation based on the computed orbit gives a mass about twice that of the Sun, a not very discordant result, as, according to Professor Pickering, there is something " peculiar" about the star's spectrum which may imply that its density and intrinsic brightness are perhaps somewhat different from that of the Sun. Compared, however, with some faint stars which show a relative proximity to our system, the Sun will contrast very favourably in size, or at least in bright- ness. A star of about the seventh magnitude in the constellation Ursa Major, numbered 21,185 m Lalande's catalogue, has been found by Winnecke to THE SUN AMONG HIS PEERS. 79 have a parallax of about half a second of arc. At the distance indicated by this comparatively large parallax the Sun would shine as a star of about 2\ magnitude, which would make it about fifty times brighter than Lalande's star. Another small star in the same constellation, number 21,258 of Lalande's catalogue, although of only 8-J magnitude, yielded to Auwers a parallax of 0*262 of a second, which may be considered as a comparatively large one. At the distance indicated, the Sun would be reduced to a fourth magnitude star, denoting that its brilliancy is about 63 times greater than Lalande's star. Two small stars of the ninth magnitude, numbered 11,677 and 17,415 in the catalogue of CEltzen and Argelander, have been found to show a similar distance, the Sun being reduced to about the fourth magnitude in both cases. Here we have a difference of five magnitudes, which implies that the Sun is a hundred times brighter than these faint, although comparatively near stars. We may therefore conclude that while some of the brighter stars are probably vastly larger than our Sun, others are almost certainly much smaller. The larger stars, overcoming as they do the dwindling effect of vast distance by their stupendous size, may possibly form exceptions to the general rule of stellar mass, and those faint stars which are at a measurable distance from the Earth, showing by their feeble light and comparative proximity that they are really as So THE WORLDS OF SPACE. well as apparently small, may also form exceptions in the opposite direction. The conclusion then seems probable that the Sun is an average-sized star, neither an exceptionally large nor an exceptionally small member of the vast and varied sidereal system which forms our visible universe. VIII. REVOLVING SUNS. SOME stars seen with the naked eye seem very close together. These, although not regarded by astrono- mers as double stars, very much resemble real double stars as seen in a telescope. Of these "naked eye doubles" may be mentioned Mizar and Alcor, a Capriconi, Tauri (in the Hyades), o Cygni, etc. But the distance which separates even the closest of these is very considerable when compared with the objects revealed by the telescope. And even among those included in double star catalogues, the distance varies considerably, from doubles which can be easily seen with telescopes of two or three inches in diameter, to those excessively close objects which require the highest powers of the largest telescopes to show them as anything but single stars. The term " double " star seems to have been first used by Ptolemy, who applied the Greek word diplons or " double " to v Sagittarii, which consists of two stars of the fifth magnitude close together, as seen with the naked eye, 82 THE WORLDS OF SPACE. The first double star discovered with the telescope seems to have been Orionis the middle star of the " sword " in which Huygens is said to have seen four stars in 1656, and y Arietis the faintest of the three well-known stars in the " head of Aries " which was first seen double by Hooke in 1664, while observing the comet of that year. " I took notice," he says, " that it consisted of two small stars very near to- gether ; a like instance of which I have not else met with in all the heavens." This is an excellent object for a small telescope, as the components are nearly equal, and separated by nearly 8" of arc. The duplicity of a Centauri is said to have been discovered by Richaud in 1690. Bradley divided y Virginis in 1718, and Castor in the same year. The appearance of these double stars naturally suggests the idea that they are comparatively close together, and at about the same distance from the Earth. It is of course possible that one of the stars might lie far beyond the other in space, thus forming what is called an " optical double." This may probably be the case when one is much brighter than the other, and the apparent distance between them is considerable. But in those double stars of which the components are nearly equal and very close, it seems more probable that they lie at nearly the same distance from the Earth. In a paper read before the Royal Society in the year 1767, by the Rev. John Michell, he showed that the mathematical REVOLVING SUNS. 83 probability of even the two stars of /3 Capricorn i being merely placed together by the effect of chance was only ^ or in other words the odds were 80 to I against the accidental proximity of these two stars. The chances against accidental arrangement are of course much increased in the case of telescopic doubles. For example, the bright star Castor consists of two nearly equal stars separated by a distance less than - a Vth of that which divides the components of ft Capricorni, and there are of course many doubles very much closer than Castor. In the case of three or more stars lying close together, the chances are of course still further increased. In the case of the Pleiades, Michell found that the odds were no less than 500,000 to i against the accidental grouping of the six brightest stars of this famous cluster. Now that photography has shown that the Pleiades contain over 2000 stars, the probability rises to an absolute certainty that they are physically connected. From these considerations Michell concluded that in all probability many of the double stars form systems connected by the laws of universal gravitation. He did not, however, put this very plausible hypothesis to the test of observation, and it was reserved for Sir William Herschel to discover towards the close of the eighteenth century the actual existence of revolving stars. It does not appear, however, that Herschel was acting on Michell's suggestion when he made the discovery with which his name is usually 84 THE WORLDS OF SPACE. associated, and to Michell seems clearly due the discovery from abstract reasoning of the probable existence of these stellar systems. Herschel's observ- ations were made with a view to determine the distance of certain double stars from the Earth. Supposing one of the stars to be much nearer to the Earth than the other, there would result an annual swaying to and fro of the components, due to the Earth's revolution round the Sun. Instead of this apparent annual motion, Herschel found, after twenty- five years of observation, that in many cases there was a progressive motion of one of the components constantly in the same direction, thus indicating that the stars were actually revolving round each other. Here was a remarkable discovery, one of the most wonderful of modern times the existence of revolv- ing suns. As Herschel himself expressed it, "he went out, like Saul, to seek his father's asses, and he found a kingdom." For some twenty years after Herschel's interesting discovery announced in 1803 little seems to have been done in the observation of double stars. This apparent neglect was princi- pally due to the fact that telescopes sufficiently powerful to deal satisfactorily with the closer double stars had not yet been constructed. The work was however soon taken up by Sir John Herschel in England and Struve in Russia, and these famous astronomers added greatly to our knowledge of the subject. REVOLVING SUNS. 85 Between the years 1830 and 1868, valuable work was done in this interesting branch of astronomical research by Dawes, Dembowski, Jacob, Madler, Powell, Secchi, Smyth, and others ; and in more recent years by Burnham, Doberck, Englemann, Glasenapp, Schia- parelli, Tarrant, etc. The number of known double stars has been largely increased by the labours of Mr. Burnham, who has discovered over 1000 pairs, most of them very close, and many of them certainly in orbital motion. The first attempt at calculating the orbit of a binary, or revolving double star, was made by Savary in the year 1830. The star selected was f Ursre Majoris. This remarkable pair of suns was dis- covered by Sir William Herschel in 1780. Savary found a period of revolution of 58^- years, but a slightly longer period of 6o|, computed by Duner in 1876, is probably nearer the truth. The companion star has therefore made one complete revolution round its primary since its discovery, and is now far advanced in a second revolution. Although the components are not at present at their greatest distance apart, they are yet within the range of telescopes of moderate size. 1 The next binary star which seems to have been attacked by the astronomical computer was 70 Ophiuchi, for which Encke computed, in 1830, a 1 A diagram of the orbit of this remarkable pair will be found in Knowledge for August 1890, 86 THE WORLDS OF SPACE. period of 79 years, and Sir John Hcrschel, in 1833, 80 years. Recent calculations, however, make the period somewhat longer, an orbit computed by the present writer in 1888 giving a period of about 88 years. An orbit by Mr. Mann, published in 1890, also makes the period 88 years, a period also found by Mr. Burnham in 1893. More than a complete revolution has been performed since its discovery by Sir William Herschel in 1779. An orbit for the famous pair y Virginis was computed by Sir John Herschel in 1833. He found periods of 513 and 629 years, but more recent calculations have reduced this period to about 180 years. In the year 1836 the distance between the components of this remarkable pair became so small that only the largest telescopes could show any sign of the star being double. The stars have now opened out again, and are at present visible with even a small telescope. 'The Story of Gamma Virginis' has been well told by Admiral Smyth in his admirable Celestial Cycle (vol. ii.), and forms a most interesting chapter in the history of astronomy. An orbit for Castor was next computed by Sir John Herschel, in 1833. He made the period about 25 2 J years, but recent calculations appear to indicate a much longer period. Orbits were also computed in 1833 by Sir John Herschel for f Urs?e Majoris, f Bootis, and ?? and Persei (54 Andromedse), was described by Schweizer as " etoile rouge presentant un petit disque " in January 1843; Birmingham noted it as "light red" in December 1875; Copeland "deep red" in January 1876; and Dreyer "reddish" in September 1878 ; but Espin, in November and December 1887, found it "certainly not red, and nothing peculiar in the star's appearance." It might be expected that these curious changes of colour, if real, would be accompanied by corresponding changes in the star's spectrum. Such may be the case, and observations in this direction would probably lead to some interesting results. There seems to be some law governing the dis- tribution of the coloured stars. The white stars appear to be most numerous, as a rule, in those constellations where bright stars are most abundant, for instance in Orion, Cassiopeia, and Lyra ; yellow and orange stars in large and ill-defined constellations l6o THE WORLDS OF SPACE. such as Cetus, Pisces, Hydra, Virgo, etc. The very reddish stars are most numerous in or near the Milky Way, and one portion of the Galaxy between Aquila, Lyra, and Cygnus was termed by Birming- ham " the red region in Cygnus." XVI. SIRIUS AND ITS SYSTEM. SlRiUS, or the Dog Star, is the brightest star in the heavens, and from its superior brilliancy has been termed " the monarch of the skies/' Measures of its light show that it is about two magnitudes, or over six times brighter than an average first magnitude star like Altair or Spica, and about equal in lustre to three stars like Vega or Capella. Sir John Herschel found the light of Siritis equal to 324 times the light of a star of the sixth magnitude, about the faintest visible to average eyesight. But it is probably over 600 times brighter than a sixth magnitude star. It has been seen in daylight with a telescope of only half an inch in aperture. Some observers have even seen it with the naked eye in sunshine, and it has been observed to cast a shadow like Venus when at its brightest. The origin of the name Sirius is somewhat doubtful. It may possibly be derived from the Sanscrit word surya, the sun. Professor Max Miiller thinks that the M 1 62 THE WORLDS OF SPACE. Greek word seirios comes from the Sanscrit svar or suonasirau. Sirius is first mentioned as a star by Hesiod, who connects it with the dog days. These, according to Theon of Alexandria, commenced 20 days before Sirius rose with the Sun, and ended 20 days after that date. These so-called dog days com- mence on July 3, and end on August 1 1 ; but, owing to the precession of the equinoxes, Sirius does not now rise with the sun or heliacally, as it is termed until August 25, or 14 days after the dog days have ended. The fancied connection of Sirius with the 40 days of summer heat has, therefore, no longer any existence, and must like many such ideas be consigned to " the myths of an uncritical period." Sirius was worshipped by the ancient Egyptians under the names of Sothis (Horus), Anubis, and Thoth, and represented as a man with the head of a dog. Some identify it with the Mazzaroth of Job. It was also supposed to represent Orion's hound, and it may perhaps be identical with the Cerberus of the Greeks. It seems to be a popular idea that Sirius, now of a brilliant white colour, was a red star in ancient times. But such a remarkable change of hue is not well estab- lished. It seems more probable that the idea of change is due to the mistranslation of a word applied to the star by the ancient writers, a word which probably referred to its brightness rather than its colour. Pr t SIRIUS AND ITS SYSTEM. 163 T. J. J. See has, however, recently collected strong evidence from the classical writers to show that Sirius was really a red star in ancient times. Such a change would, of course, be most interesting and remarkable, indicating, as it would, some wonderful change in the star's chemical constitution. Like many other stars, Sirius has a considerable " proper motion " across the face of the sky, amounting to about 1*3" of arc per annum. Some irregularities in this proper motion led the astronomers Bessel, Peters, and Safford to the conclusion that the motion of Sirius was disturbed by the attraction of an in- visible close companion revolving round it. From the recorded observations Peters computed an orbit for the supposed companion, and found a period of about 50 years. Safford also investigated the pro- blem, and announced in 1861 the probable position of the invisible companion. About four months after the publication of Safford's results, Mr. Alvan Clark, the famous American optician, observing with a tele- scope of 1 8 inches aperture, detected a small star near Sirius, the position of which agreed closely with that of Safford's hypothetical companion. Here was a case somewhat similar to the discovery of the planet Neptune the prediction, by mathematical analysis, of the existence of a celestial body previously un- known to astronomers. Numerous observations of this small star have been made since its discovery, 164 THE WORLDS OF SPACE. and there is now no doubt that it is revolving round its brilliant primary. That the observed irregularities in the proper motion of Sirius are wholly due to the influence of this companion seems, however, to be still an open question. Several orbits have been computed, most of which assign a period of 49 or 50 years ; but an orbit recently computed by the present writer gives a period of about 58 J years, and Howard finds a period of 57 years. Burnham, however, thinks that 53 years is probably nearer the truth. As the companion has now approached Sirius so closely as to be invisible with even the giant telescope of the Lick Observatory, some years must elapse before the exact length of the period can be definitely settled. The great brilliancy of Sirius has naturally suggested proximity to the Earth, and modern measures of its distance have confirmed the accuracy of this idea. The most reliable determinations of its parallax (or the angle subtended by the radius of the Earth's orbit at the place of the star) make it about four-tenths of a second of arc, and places it about fourth in order of distance from the Earth. 1 Assuming a parallax of 0*39 of a second (about a mean of the results found by 1 The three nearest stars are : a Centauri (parallax 076 of a second), 61 Cygni (0-45"), and Lalande 21,185, for which Kapteyn found a parallax of 0-434", and Winnecke 0*5". For the star r) Herculis a parallax of 0*40" was found by Belopolsky and Wagner ; but this does not seem to have been confirmed by any other astronomer. SIRIUS AND ITS SYSTEM. 165 Drs. Elkin and Gill), the distance of Sirius would be 528,884 times the Sun's distance from the Earth, a distance which light would take about 8| years to traverse. Knowing the distance of Sirius from the Earth, and its annual proper motion, it is easy to calculate its actual velocity in a direction at right angles to the line of sight. This comes out about ten miles a second. The spectroscope shows that Sirius has also a motion in the line of sight, and hence its real velocity through space must be greater than that indicated by its proper motion. In the year 1864 observations by Dr. Huggins showed that Sirius was receding from the Earth at the rate of 29 miles a second. Some years afterwards careful measures of the star's spectrum showed that this motion had ceased ; subsequent measures showed that the motion was reversed, and recent observations by Dr. Vogel indicate unmistakably that the motion has now been changed into a motion of approach ! It seems difficult to understand how this curious change in the direction of the star's motion can be accounted for otherwise than by orbital movement ; in the same way that the planet Venus is sometimes approaching the Earth and sometimes receding from it, owing to its orbital motion round the Sun. The motion may possibly be due to the existence of some invisible close companion. I 66 THE WORLDS OF SPACE. Placed at the distance of Sirius, the Sun would, I find, be reduced to a star of only the third magnitude, or about four magnitudes brighter than Sirius appears to us. This indicates that Sirius is about 40 times brighter than the Sun would be in the same position, and would imply that Sirius is a far more massive sun than ours. If we assume the same intrinsic brilliancy of surface and the same density for both bodies, the above "result would make the diameter of Sirius 632 times the Sun's diameter, and its mass no less than 253 times the mass of the Sun. As, how- ever, the intrinsic brightness of the surface of Sirius and its density, or specific gravity, may differ widely from those of the Sun, these calculations are of course open to much uncertainty. The light of Sirius, analyzed by the spectroscope, differs considerably from the solar light, and the strong development of the hydrogen lines in the star's spectrum denotes that Sirius is, in its chemical constitution, not com- parable with our Sun. It may possibly be very much hotter, and therefore smaller in diameter and mass than the figures given above would indicate. For- tunately we can find the mass of a binary or revolving double star by another and more certain method. Knowing the orbit of the star and its distance from the Earth, we can calculate the combined mass of the components in terms of the Sun's mass. Making the necessary computations for Sirius, I find that the SIR1US AND ITS SYSTEM. 167 combined mass of Sirius and its companion is a little over three times the mass of the Sun, and the mean distance between them 22 times the Sun's distance from the Earth, or a little more than the distance of the planet Uranus from the Sun. This result recently confirmed by Dr. Auwer's calcula- tions would imply that Sirius is intrinsically a much brighter sun surface for surface than ours, and that " the monarch of the skies " is a " giant " only in ap- pearance ; the greater brightness of its surface and its comparative proximity to the Earth accounting for its great apparent brilliancy. The companion of Sirius has been estimated as of the tenth magnitude. This would imply that the light of Sirius is about 25,000 times the light of the small star. If, therefore, the two bodies were of the same density and intrinsic brightness, the mass of Sirius would be about 4,000,000 times as great as the mass of the companion. But Dr. Auwers con- cludes, from his researches on the proper motion of Sirius, that the companion is about one-half the mass of the primary, and equal in mass to our Sun ! It must, therefore, be nearly a dark body. It has been suggested that the companion may possibly shine by reflected light from Sirius, in the same way that the planets of the Solar System shine by reflected light from the Sun. Some calculations which I have recently made show, however, that this hypothesis is 1 68 THE WORLDS OF SPACE. wholly untenable. 1 Assuming, with Auwers, that the mass and diameter of the companion are equal to those of the Sun, I find that the companion would, if illuminated solely by reflected light from Sirius, shine as a star of only i6J magnitude. A star of this magnitude about the faintest visible in the great Lick telescope placed close to a brilliant star like Sirius would, even when most favourably situated, be utterly invisible in our largest telescopes. If its mass is much less than one-half that of Sirius as its faint- ness would seem to suggest it is possibly a com- paratively small body, and the reflected light from its primary would be proportionately less. It seems clear, therefore, that the companion must shine with some inherent light of its own, otherwise it could not possibly be so bright as the tenth magnitude. It is probably a sun of small luminosity revolving round Sirius in the same way that the companions to other binary stars revolve round their primary. The dis- parity in brightness is, however, remarkable, no other binary star showing so great a difference in the brilliancy of the components. As I have said above, the Sun, if placed at the distance of Sirius, would shine as a star of the third magnitude. There is, therefore, a difference of seven stellar magnitudes between the light of the Sun and 1 Journal of the British Astronomical Association, March 1891. SIRIUS AND ITS SYSTEM. 169 that of the Sirian satellite. This implies that the light emitted by the Sun is 631 times greater than that radiated by the companion of Sirius. If of the same intrinsic brightness of surface, the latter would, therefore, have a diameter about ^ T th of the Sun's diameter, or 34,000 miles. But if of the same mass as the Sun, its density with this small diameter would be enormous in fact, vastly greater than we can imagine possible for any body large or small. Indeed, if we suppose its diameter to be one-half that of the Sun, its density would be 11*52 (1*44 X 8), or about equal in density to lead, and it seems very improbable that a self-luminous body could have so high a density as this. We must conclude, therefore, that the satellite of Sirius is a comparatively large body having a small intrinsic brilliancy of surface possibly a cooling body verging towards the utter extinction of its light. If this be so, it will probably, in the course of ages, disappear altogether from telescopic vision, and its continued existence will only be known by its influence on the motion of Sirius. If there are any planets revolving round Sirius they will probably remain for ever unknown to us. A planet comparable with Jupiter in size would be utterly invisible in the giant telescope of the Lick Observatory, or even with an instrument very much larger. I am disposed, however, to think that these binary stars may perhaps form exceptions to the I7O THE WORLDS OF SPACE, general rule of stellar systems, and that single stars, like our Sun, more probably form the centres of planetary systems like our own. Or possibly the reverse of this may be true, the single stars forming the exceptions, and binary stars the rule. In either case we may conclude, I think, judging from the analogy of our Sun, that single stars are more likely to have planets revolving round them. XVII. DARKENINGS OF THE SUN. THERE are many cases recorded in history of the Sun having been remarkably darkened, and the daylight obscured for periods of varying duration. Calculation shows that some of these were undoubt- edly due to total eclipses of the Sun, but others cannot be so easily explained. Plutarch speaks of the paleness of the Sun during the year 44 B.C. This was about the time of the assassination of Julius Caesar, and calculation proves that no total eclipse of the Sun occurred in that year. The darkness recorded at the Crucifixion of our Saviour was certainly not due to an eclipse of the Sun, as has been suggested by some ignorant sceptics. For the darkness occurred at the Paschalfut/moon, and lasted three hours, facts quite irreconcilable with the conditions of a solar eclipse. The darkness recorded in the Gospels is also mentioned by contemporary historians. In the year A.D. 192 it is said that "stars were seen 1/2 THE WORLDS OF SPACE. in the daytime," but there was certainly no total eclipse of the Sun that year. Again, it is recorded that when the famous Alaric appeared before the walls of Rome (A.D. 409) a dark- ness set in so great that stars were seen in the daytime. Shortly before the capture of the city (August 410) there was a partial eclipse of the Sun visible at Rome (June 18), but the central line of the eclipse which was an annular one passed consider- ably south of Rome. In the year 536 A.D. the Sun's light is said to have been greatly diminished, and to have remained so for about fourteen months ! In 626 and 627 the Sun is said to have lost half its light ! There seem to have been no great solar eclipses in either of these years. According to Schnurrer, a remarkable Sun darken- ing occurred in September 1091, which lasted three hours. There was no total eclipse of the Sun at this time ; but a darkening recorded by the same writer in June 1191 may be explained by a great solar eclipse mentioned by English writers which took place on June 23 of that year. The eclipse was, however, only an annular one, so that the darkness could not have been great. This fact tends to show that the descriptions given of these phenomena by the early writers are probably much exaggerated. In the Chinese Annals " a great diminution of light " is recorded on July 8, 1103, which cannot be accounted for by an eclipse. The same remark DARKENINGS OF THE SUN. 173 applies to a darkening on February 12, 1106, men- tioned by Erman as having been accompanied by meteors. The allusion to meteors is very significant, suggesting the probable interposition of a meteoric swarm between the Sun and Earth. On February 4 of the same year a great comet was seen close to the Sun in full daylight. It has been thought probable that this comet is identical with the great " September comet" of 1882. Possibly a swarm of meteors was travelling in the wake of the comet of 1106, and passed between the Sun and Earth, or perhaps actually collided with the Earth as the Leonids do. A Spanish writer relates that on the last day of February 1206 there was total darkness for six hours. With reference to this startling statement, it seems worth remarking that an eclipse of the Sun is recorded by several writers on February 28, 1207, so that if we suppose the writer to have made a mistake of one year, and if for " six hours " we read six minutes, the occurrence might be explained. In April 1547 the Sun's light is said to have been so diminished that for three days stars were seen in the daytime, the Sun appearing " as though suffused with blood." There was no total eclipse in this year. It seems worthy of notice that some of the dates mentioned above fall close to calculated epochs of Sun-spot maxima. Although in recent years there has been no extraordinary development of Sun-spots at the epoch of maxima, it is not altogether impossible 1/4 THE WORLDS OF SPACE. that in former times these spots may have occasion- ally increased to such an extent, both in number and size, as to have perceptibly darkened the Sun's light. A more probable explanation, however, seems to be the passing of a meteoric or nebulous cloud between the Sun and Earth. The most recent instance of Sun-darkening recorded in this country occurred on May 22, 1870, when the Sun's light was observed to be considerably reduced in a cloudless sky in the west of Ireland ; at Greenwich on the 23rd, and on the same day, but at a later hour, in north-eastern France " a progressive manifestation that seems to accord well with the hypothesis of moving nebulous matter." A similar phenomenon was observed in New England on September 6, 1881. XVIII. THE NUMBER AND DISTANCE OF THE VISIBLE STARS. THAT the visible stars are not uniformly scattered through space, and are not of uniform size and intrinsic brightness, is clearly shown by modern researches. Measures of stellar parallax show that some small stars (that is, faint stars) are actually nearer to our system than many of the brighter stars, while the period of revolution of some binary stars show that their mass is relatively small compared with the brilliancy of their luminous surfaces. We may, however, perhaps assume that the stars, out to some limited distance in space, are scattered with some rough approach to uniformity. We can at least calculate the average distance between the neighbouring stars, which would give a certain number of stars in a sphere of given radius. This is easily done by supposing the stars placed at the angular points of a tetrahedron. A tetrahedron is a solid figure bounded by four equal surfaces, each surface being an equilateral triangle. It is clear that THE WORLDS OF SPACE. in such a solid, each of the angular points is equi- ithe other three angular points of the ?, ~nvi tilat stars so placed in space would be " uniformly distributed " through the space containing them. With this arrangement the number of stars con- tained in a given sphere may be easily calculated, assuming a certain distance between each pair of stars ; but the calculation will only apply to a sphere of a radius very large in comparison with the distance between the stars distributed through it. If we suppose the distance between the stars equal to the radius of the sphere, the calculation gives, I find, 35 equi-distant stars in the sphere. This number is evidently too great, as the number of stars which can be placed on the surface of a sphere of given radius, equi-distant from each other and from the centre of the sphere, is only 12. The difference is clearly due to the fact, that in this case the volume of the tetrahedron is so large in proportion to the volume of the sphere that the latter cannot be accurately divided into tetrahedrons. In the case, however, of a sphere whose volume is very large in proportion to the volume of the tetrahedron formed by four ad- jacent stars, the calculation will be approximately correct. Let us call the distance between two adjacent stars the unit distance. Now considering a Centauri, for which the largest parallax has been found (about 076"), it is obvious NUMBER AND DISTANCE OF VISIBLE STARS. 1 77 at once that this star cannot be at the "unit distance" from the Sun. For if a Centauri was at the "unit distance," we might expect to find some ten or eleven other stars at a similar distance from the Earth. Such is, however, probably not the case, and we may therefore conclude that this star is comparatively near our system, and forms an exception to the general rule of stellar distance. To make this point clearer, let us see what number of stars should be visible to the naked eye say to the sixth magnitude inclusive on the assumption that the distance between the Sun and a Centauri forms the " unit distance " between two stars of the visible sidereal system, or at least that portion of the system which is visible without a telescope. To make this calculation, it will of course be necessary to assume some average distance for stars of the sixth magnitude based on actual measurement. Now Peters found an average parallax of 0*102" for stars of the first magnitude, Gylden found 0*083", and Elkin 0*089". These results are fairly accordant, and we may assume the mean of these values, or 0*09", as the mean parallax of an average star of the first magnitude. With a "light ratio" of 2-512, the light of a first magnitude star is 100 times the light of a sixth, and hence as light varies inversely as the square of the distance the distance of an average sixth magnitude star would be ten times that of a first. Its parallax N i;8 THE WORLDS OF SPACE. would therefore be 0*009". Hence the radius of the sphere containing all stars to the sixth magnitude inclusive would be 076 divided by 0*009, or $4'4 times the distance of a Centauri. From this I find that the number of stars contained in the sphere would be 21,372,000, a number enormously greater than the known number of stars to the sixth magnitude. This result shows that, on the hypothesis of uniform distribution, a Centauri cannot lie at the " unit distance " from the Sun, and that it is probably an exceptionally near star. The same may be said of 6 1 Cygni, Sirius, and other stars with a large parallax. But the above result also leads us to doubt whether the mean parallax of sixth magnitude stars is so small as O'OO9". I find that the Sun placed at the distance indicated by this parallax would shine only as a star of the eleventh magnitude l ; that is, a sixth magnitude star would be five magnitudes, or loo times brighter than the Sun placed in the same position. If of the same intrinsic brilliancy of surface, this would imply that an average sixth magnitude star has ten times the diameter of the Sun, and therefore 1000 times its volume ! Some sixth magnitude stars may possibly exceed our Sun in size, but that the average volume of these small stars is IOOO times that of the Sun seems wholly improbable. Certainly the calculated masses of those 1 See chapter on * The Sun among his Peers*' NUMBER AND DISTANCE OF VISIBLE STARS. I/Q binary stars of which the distance from the Earth has been determined, do not give any grounds for supposing that such enormous bodies exist among stars of the sixth magnitude. Assuming, however, that the parallax of a first magnitude star is 0*09", and that of a sixth magni- tude star is one-tenth of this, or oxxx)", let us see what number of stars should be visible to the sixth magnitude. As already stated, the number of equi- distant stars which can be placed on the surface of a sphere of unit radius is 12. Hence on the surface of a sphere of double this radius, four times the number, or 48 equi-distant stars may be placed ; on a sphere of three times the radius, nine times the number, and so on. Now the sum of ten terms of this series I 2 , 2 2 , 3 2 , etc., is 385, and as the twelve stars which may be placed on the first sphere nearly represent the number of stars of the first magnitude and brighter visible in both hemispheres, we have the total number of stars to the sixth magnitude inclusive 385 x 12, or 4620. Now the number of stars to the sixth magnitude in both hemispheres, as observed by Heis and Gould, is 4181, and the number contained in the Harvard PJiotoinetry^ and the Uranometria Argentina is 3735, so that the number of stars computed on the above principle does not differ widely from the number actually observed. Let us see now what the unit parallax would be ISO THE WORLDS OF SPACE. for the observed number of stars to the sixth magni- tude, assuming a parallax of 0*009" f r stars of this magnitude. Taking the number as 3735, I find by the tetrahedron hypothesis that the parallax of the star at " unit distance," that is the mean parallax ot the nearest stars to the Earth, would be 0*042". Ex- cluding stars with a large parallax, this may not be far from the truth. I find that in 31 binary stars brighter than the sixth magnitude (and for which a parallax has not yet been determined), the average " hypothetical parallax " or the parallax on the assumption that the mass of the system is equal to the mass of the Sun is 0*068". If we assume the mass of each of these systems to be, on an average, twice the Sun's mass, we must divide this by the cube root of 2. This gives for the average parallax 0*054", which does not differ widely from the unit parallax found above for stars of the sixth magnitude. But we are still confronted with the difficulty that with a parallax of only oxx>9", stars of the sixth magnitude would be on an average considerably larger than the Sun. The same remark applies to stars of the sixteenth magnitude, for which the parallax would be only OXDOOOQ". Placed at this vast distance, the Sun would, I find, be reduced to a star of magnitude 21*3, and would, therefore, be utterly invisible in the largest telescopes yet con- structed. It would be over 100 times fainter than a star of the sixteenth magnitude ! NUMF.ER AND DISTANCE OF VISIBLE STARS. l8l To reduce the Sun to a star of the sixth magnitude, it should be placed at a distance corresponding to a parallax of 0*1". Unless, therefore, stars of the sixth magnitude are, on the average, considerably larger than our Sun, we seem justified in thinking that their average parallax is not less than one-tenth of a second. But, as has been stated, this is about the average parallax of stars of the first magnitude, and it is difficult to believe that these bright stars are as far from the Earth as the faint stars which lie near the limit of naked-eye vision. There seems, however, no escape from the conclusion that sixth magnitude stars are probably nearer to us than their brightness might lead us to suppose, and to explain the difficulty with reference to the brighter stars, we may perhaps assume that their brilliancy is due rather to their great size than proximity to our system. From the small parallax found for Arcturus, Vega, Capella, Canopus, and other bright stars, we have good reason to think that these stars are vastly larger than our Sun. 1 Spectroscopic observations of f Ursre Majoris indicate that this second magnitude star has a mass about 40 times the mass of the Sun, and possibly other bright stars may have similarly large masses. Sirius and a Centauri, however, form notable excep- tions to this rule. Assuming an average parallax of one-tenth of a second for stars of the sixth magnitude, the parallax 1 See chapter on ' The Suni-among his Peers,' 1 82 THE WORLDS OF SPACE. of a star at the " unit distance " from the Sun would be 0*47". This is about the parallax found for 61 Cygni, and does not much exceed that of Sirius. There are several other stars with a parallax of somewhat similar amount, and possibly there may be others hitherto undetected. With a parallax of OT" for a sixth magnitude star, the parallax of an eleventh magnitude would be 0*01", and that of a sixteenth magnitude 0*001". Now with the unit distance corresponding to a parallax of 0*47", I find that the number of equi-distant stars contained in a sphere of radius equal to the distance of a sixteenth magnitude star would be 3,690,700,000, a number about 36 times greater than the number of the visible stars, generally assumed at 100,000,000. According to Dr. Gould's formula, the number of stars to the sixteenth magnitude would be 3,024,057,632, or about 30 times the number actually visible. Probably, however, we are not justified in assuming a uniform distribution of stars to the sixteenth magnitude, most of these faint stars belonging to the Milky Way. Professor Celoria found that, near the pole of the Galaxy, a small telescope, which showed stars to only the eleventh magnitude, revealed as many stars as Herschel's large gauging telescope of iS'8 inches aperture. Here, therefore, we seem to have the extension of our sidereal system limited to the distance of eleventh magnitude stars. Let us now assume a uniform distribution of stars to the NUMF.ER AND DISTANCE OF VISIBLE STARS. 183 eleventh magnitude. With a parallax of O'Ol", and a unit parallax of 0*47", I find the number of stars 3,690,700. The number by Gould's formula is 3,283,876. Both results are largely in excess of the number actually visible, and show, I think, that there is probably a "thinning out" of the stars before we reach the eleventh magnitude distance, at least in extra Galactic regions. If we suppose that of the 100,000,000 of visible stars, 50,000,000 are scattered uniformly through a sphere having a radius equal to the distance assumed for stars of the eleventh magni- tudethe remaining 50,000,000 being included in the Milky Way we have an average "unit parallax" (that is, the parallax of a star as seen from its nearest neighbour) for these 50,000,000, of about o'li", which seems to indicate a "thinning out" of the stars towards the boundaries of our visible sidereal system. If this be so, we may conclude that the stars with a larger parallax than O'li" are exceptions to the general rule of stellar distribution, and form, perhaps, comparatively near neighbours of our Sun. These near stars, seen from the outskirts of the visible universe, may perhaps form a small open cluster. Thus the parallax of a Centauri being 076", the Sun and a Centauri, seen from a sixteenth magnitude star equally distant from both would appear as two faint stars about 4^' of arc apart. The above results are of course based on the assumption that the faint telescopic stars lie at a 1 84 THE WORLDS OF SPACE. distance indicated by their brightness. Such, how- ever, may not be the case. Many of these small stars may be in reality absolutely small. The apparently close connection between bright and faint stars, as shown by photographs of the Milky Way near a Cygni and a Crucis, suggests that bright naked-eye stars and faint telescopic objects may, in some cases at least, lie in the same region of space. If this be so, the difference in the size of these distant suns must be enormous, and would lead to the conclusion that possibly the Milky Way may not lie so far from our system as has been generally supposed. If, as has been suggested, there is any extinction of light in the ether which is, however, very doubtful the faintest stars cannot be placed at a distance corre- sponding to their relative brightness. XIX. SWARMS OF SUNS. AMONG the so-called nebula are many objects which, when examined with telescopes of adequate power, are seen to be resolved into myriads of small stars. Their comparative isolation from surrounding objects impresses us forcibly with the idea that they form, as it were, families of stars connected by some physical bond of union. Of these clusters, as they are called, we have naked-eye examples in the Pleiades, and the "Bee Hive" in Cancer. Others may be partially seen with a good opera-glass or binocular, but most of them require telescopes of considerable power to view them to advantage. They are of various forms and of all degrees of condensation. Some are comparatively large and irregular, others small and compressed, with the component stars densely crowded. Many are of such uniform shape as to have received the name of globular clusters. These have been aptly termed " balls of stars," and are among the most interesting objects in the stellar 1 86 THE WORLDS OF SPACE. heavens. The most remarkable example of this class visible in the Northern Hemisphere is that known as 13 Messier. It lies between the tolerably bright stars Zeta and Eta Herculis, nearer the latter star. It may be seen with an opera-glass as a hazy-looking star of about the sixth magnitude, with a star on each side of it. Examined with a powerful telescope it is resolved into numerous small stars. Sir William Herschel estimated them at 14,000, but the real number is probably much less. Assuming the average magnitude of the components at twelve and a half, I find that an aggregation of 14,000 stars of this brightness would shine as a star of about the second magnitude, or a little fainter. Examining this object with his giant telescope, Lord Rosse noticed three dark rifts radiating from the centre. These were afterwards seen by Buffham, with a 9-inch reflector, and also by Webb. They were also observed at the Ann Arbor Observatory (U. S. A.) in April 1887, by Professor Harrington and Mr. Schaeberle, using telescopes of 6 and 12 inches aperture. Professor Harrington, comparing his draw- ing with that of Lord Rosse, thinks that the rifts "have shifted their position slightly in the 50 or more years which have elapsed since the first drawing was made." This seems, however, very improbable. The suspected change may be simply due to difference in the methods of delineation, and in the relative sharpness of the observer's eyesight. This cluster SWARMS OF SUNS. iS/ has been successfully photographed at the Paris Observatory, and by Mr. Roberts at Liverpool. In these photographs the dark rifts arc traceable to some extent, but owing perhaps to over-exposure of the central portion of the cluster they are not so distinct as in the drawings referred to. Examined with the spectroscope, Dr. Huggins finds the spectrum con- tinuous, but deficient at the red end, like the great nebula in Andromeda. Mrs. Huggins has, however, pointed out that this apparent suppression of red rays is simply due to the faintness of light in these objects. 1 Spectroscopic evidence is, however, hardly necessary to prove that the Hercules cluster consists of small stars, as these are distinctly seen as points of light with telescopes of moderate power, and with the great Lick telescope the component stars are visible even in the central portion of the cluster. . Another object of the globular class, but less resolvable, is that known as 92 Messier, which lies between the stars Eta and Iota, in Hercules, nearer the latter. Sir William Herschel's telescopes showed it as seven or eight minutes of arc in diameter. It is considerably brighter at the centre. The larger components are easily visible in moderate- sized telescopes, but even Lord Rosse's giant instru- ment failed to resolve the central blaze. There is no doubt, however, that it consists wholly of small stars, as the unerring eye of the spectroscope shows a stellar 1 The Observatory, December 1890. 1 88 THE WORLDS OF SPACE, spectrum, similar to that of the neighbouring 13 Messier. Another fine example of the globular class is 5 Messier, which lies closely north, preceding the fifth magnitude star, 5 Serpentis. It is considerably compressed at the centre. Sir W. Herschel counted 200 stars, but failed to resolve the central nebulosity. Messier, its discoverer, found it visible with a telescope only one foot long. Another fine object is 3 Messier, in Bootes. Admiral Smyth describes it as "a brilliant and beautiful globular aggregation of not less than 1000 small stars." It is beyond the power of small telescopes, but it was resolved by Buffham, even in the centre, with a Q-inch reflector. Numerous fine examples of the globular class are found in the Southern Hemisphere, which indeed seems to be richer in these marvellous objects than the northern sky. Of these the most interesting are those known as Omega Centauri, and 47 Toucani. Omega Centauri from its great apparent size about two-thirds of the Moon's diameter and its visibility to the naked eye, may perhaps be con- sidered as the most remarkable object of its kind in the heavens. It shines as a hazy star of the fourth magnitude, and I have often so seen it in the Punjab sky. Its large size and globular form are clearly visible in a binocular field-glass, but of course its component stars are far beyond the reach of such an THE STAR CLUSTER CENTAURI . From a Photograph taken by D. GILL, at the Royal Obsetvatcry, Cape of Goad Hope, on May 2$th, 1892. SWARMS OF SUNS. 189 instrument. Sir John Herschel, observing it with his large telescope at the Cape of Good Hope, found it " a truly astonishing object. All clearly resolved into stars of two magnitudes, viz. thirteen and fifteen, the larger lying in lines and ridges over the smaller ; . . . the larger form rings like lace-work on it." I have shown elsewhere 1 that if we take the average magnitude of the components at thirteen and a half the apparent brightness of the cluster would imply that it contains about 15,000 stars. Another wonderful cluster is that known as 47 Toucani, which lies close to the smaller Magellanic cloud. It is smaller in apparent size than Omega Centauri, but Dr. Gould, observing it at Cordoba, speaks of it as " one of the most impressive and perhaps the grandest of its kind in either hemi- sphere," and he estimates its magnitude at four and a half, as seen with the naked eye. It is thus de- scribed by Sir John Herschel " A most magnificent globular cluster. It fills the field with its outskirts, but within its more compressed part I can insulate a tolerably defined, circular space, of go" diameter, wherein the compression is much more decided, and the stars seem to run together, and this part, I think, has a pale pinkish or rose colour . . . 1 Planetary and Stellar Studies, p. 191. On a photograph recently taken at Arequipa, Peru, 6389 stars have been actually counted, but the enumerator, Mr. Baily, considers that it includes a much larger number. THE WORLDS OF SPACE. which contrasts, evidently, with the white light of the rest. The stars are equal, fourteen magnitude immensely numerous and compressed. . . . Conden- sation in three distinct stages. ... A stupendous object." Sir John Herschel's drawing of this cluster reminds one of a swarm of bees, and perhaps suggested to Tennyson the lines, " Clusters and beds of worlds, and bee-like swarms Of suns and starry streams." There are other interesting specimens of the globular class in the Southern Hemisphere, but not of such large apparent dimensions as those already de- scribed. Of these may be mentioned 22 Messier, which lies about midway between the stars Mu and Sigma Sagittarii. It is described by Sir John Herschel as a fine globular cluster, with stars of two magnitudes, namely, eleven or twelve, and fifteen or sixteen, the larger being visibly reddish, and he suggested that it consists of " two layers, or one shell over another." Owing to the comparative brightness of the larger components, this cluster forms a good object for small telescopes. I saw the brighter stars well with a 3-inch refractor in the Punjab sky, but, of course, the greater portion of the cluster has a nebulous appearance in a telescope of this size. Between Alpha and Beta Scorpii there is a con- densed globular cluster. With small telescopes it very much resembles a telescopic comet, but with larger SWARMS OF SUNS. 19! instruments its true character is revealed. Sir William Herschel considered it " the richest and most con- densed mass of stars in the firmament." In May 1860, a " temporary star " of the seventh magnitude suddenly appeared in the centre, almost blotting out the cluster by its superior light. The star faded away before the end of June of the same year, and has not been seen with any certainty since. It has been suggested that this temporary star lay betiveen the cluster and the Earth, but it seems to me much more probable that the outburst took place in the cluster itself, and that it was possibly caused by a collision between two of the component stars, or by a swarm of meteors rushing with a high velocity through the cluster. The beauty and sublimity of the spectacle presented by these globular clusters, when viewed with a power- ful telescope, is such as cannot be adequately de- scribed, and it has been said that when seen for the first time, " few can refrain from a shout of rapture." The component stars, although distinctly visible as points of light, defy all attempts at counting them, and seem literally innumerable. Placed like a mass of glittering diamond-dust on the dark background of the heavens, they impress us forcibly with the idea that if each of these lucid points is a sun, the thou- sands which seem massed together in so small a space must be in reality either relatively close and indi- vidually small, or else the system of suns must be 192 THE WORLDS OF SPACE. placed at a distance almost approaching the infinite The former hypothesis is perhaps the more .probable, although it is not easy to imagine, on mechanical principles, how an immense assembly of bodies filling a globular space can exist in that condition without inter- fering with each other's motions. At rest they cannot be, as their mutual attractions would soon produce a velocity in each member of the system. They must therefore be in motion, each star, perhaps, describing its own ellipse round the centre of gravity of the whole mass, which is probably situated near the centre of the sphere. The distance of these globular clusters from the Earth is, however, certainly very great. Attempts to accurately determine their position in space have not been attended with success. As the component stars are at practically the same distance from the eye, we have no comparison stars to measure from, and their exact distance therefore remains unknown. We may, however, estimate their probable distance with some show of plausibility. We may assume that the stars of the Hercules cluster would, if concentrated in a point, shine as a star of about the fourth magnitude. As the components are of the twelfth and thirteenth magnitudes, this would imply that the cluster consists of about 2500 stars. Now, assuming the average dis- tance found by Dr. Elkin for stars of the first magni* tude (about 36 years of light travel), I find that a star of the fourth magnitude would be at such a SWARMS OF SUNS. IQ3 distance from the Earth that its parallax, as it is called, would be about one-fiftieth of a second of arc, a distance which light, with its velocity of 186,000 miles a second, would take 148 years to traverse ! Now, neglecting the outliers of the cluster, we may take the apparent diameter of the more condensed part at 5' of arc (about one-sixth of the Moon's diameter). This, with the assumed distance, would denote that the real diameter of the cluster is about 15,000 times the Sun's distance from the Earth, which would give a distance between each component of about 890 times the Sun's distance, or about 29 times the distance of Neptune from the Sun. Hence, although apparently crowded together, the constituent stars may possibly be separated by immense intervals. Placed at the vast distance assumed for the cluster, our Sun would appear as a small star of between the ninth and tenth magnitudes. Each component of the cluster shines therefore with one-sixteenth of the solar light, and, if of the same density, would have one-sixty-fourth of the Sun's mass. The total mass of the cluster would therefore be equal to about 40 suns. With the data assumed, we may therefore conclude that the components of the Hercules cluster are suns of comparatively small size, separated by considerable distances, but apparently massed together by the effect of distance. Among less condensed star clusters there are many interesting objects. The Pleiades have been already O 194 THE WORLDS OF SPACE. referred to. On a photograph of this remarkable group, taken at the Paris Observatory, over 2000 stai s can be counted of all degrees of brilliancy, from those visible without optical aid, down to points of light so faint as to be invisible to the eye in the telescope with which they were photographed. Here we have a cluster of probably larger size than that in Hercules, possibly at a greater distance from the Earth, and with its larger components of considerably greater- mass than that of our Sun. Near the bright star Pollux, I see a small cluster of stars of about the seventh and eighth magnitudes, which, with a binocular field-glass, very much re- sembles the Pleiades as seen with the naked eye. A similar cluster (known as 39 Messier) may be seen near the star TT' Cygni. The well-known double cluster, x Persei, may be also seen with an opera-glass, but a telescope is necessary to show the component stars to advantage, and the larger the telescope the greater the number of faint stars visible in these wonderful objects. They have been well photographed at the Paris Observatory, and on the photograph the clusters are clearly resolved (at least on the paper print in my possession), with no trace of outstanding nebulosity, suggesting that the component stars are probably at nearly the same distance from the earth. The cluster known as 35 Messier, a little north of the star 17 Geminorum, is visible in an opera- M 37 AURIG/E 5 h 45 m + 32 31'. February %th, 1893. 90 minutes exposure. SWARMS OF SUNS. 1Q5 glass, but a small telescope is required to see the component stars. A beautiful photograph of this cluster has also been obtained at the Paris Observ- atory. A well-marked clustering tendency is visible among the brighter stars of the group, two, three, four, and sometimes five stars being grouped together in subordinate collections. Admiral Smyth says " It presents a gorgeous field of stars from the ninth to the sixteenth magnitude, but with the centre of the mass less rich than the rest. From the small stars being inclined to form curves of three or four, and often with a large one at the root of the curve, it somewhat reminds one of the bursting of a sky- rocket." This tendency to " stream " formation in the components of star clusters is also well marked in a photograph of the cluster 38 Messier (kindly sent to me by M. M. Henry, of the Paris Observatory). It was described by Webb as " a noble cluster, arranged in an oblique cross," and Smyth says " The very unusual shape of this cluster recalls the sagacity of Sir William Herschel's speculations upon the subject, and very much favours the idea of an attrac- tive power lodged in the brightest part. For although the form is not globular, it is plainly to be seen that there is a tendency towards sphericity, by the swell of the dimensions as they draw near the most lumin- ous part, denoting, as it were, a stream or tide of stars, setting towards the centre." Sir William Herschel, speaking of a compressed 196 THE WORLDS OF SPACE. cluster in Perseus, says " the large stars are arranged in lines like interwoven letters," and Webb says " it is beautifully bordered by a brighter foreshortened pentagon." Observing with a 3-inch telescope in India, I noticed a beautiful cluster of stars, about 4 north of A and v Scorpii, resembling in shape a bird's foot, with remarkable streams of stars. This cluster is visible to the naked eye as a star of about the fifth magnitude. Although these loosely-associated star clusters do not show such evidence in favour of family connection as the more closely-compacted globular clusters, still we can hardly escape from the conviction that their apparent aggregation is really due to some physical bond of union, and not merely the result of a fortuitous scattering of stars at different distances in the line of sight. XX. C;REAT NEBULAE. THOSE " dim and mysterious" objects known as nebulae present themselves under very various aspects to the inquiring eye of the astronomer. Many are small, faint, and ill-defined, even in the largest telescopes. Others are large, irregularly shaped, and comparatively bright. Some are circular or elliptical in outline ; others annular, elongated, or comet-shaped. Some show a spiral structure. Some are single ; others double and even triple. In the present paper we will consider the large and irregular nebulae. Of these, the most remarkable object in the Northern Hemisphere is the great nebula in Andro- meda, known to astronomers as 31 Messier, and sometimes described as "the queen of the nebulae." On a clear moonless night it may be just detected with the naked eye as a hazy spot of light near the 4| magnitude star v Andromedae, and even with an opera-glass it is a striking and beautiful object. It was probably known to the ancients, and it could hardly have escaped their keen eyesight in the clear Eastern skies. It was certainly seen as far back as A.D. 905. Al-Sufi, the Persian astronomer, who wrote 198 THE WORLDS OF SPACE. a description of the heavens in the middle of the tenth century, speaks of it as a familiar object in his day, and the nebula is marked on a star map made in Holland about the same period. It escaped observation, however, by Tycho Brahe and Bayer. Simon Marius described it in 1612 as resembling a candle shining through a horn, and it was seen by Bulialdus in 1664, while observing the comet of that year. Halley explained it as "nothing else but the light coming from an extraordinary great space in the ether, through which a lucid medium is diffused that shines with its own proper lustre." A small adjacent nebula, a little to the north-west of it, was discovered by Le Gentil in November 1749, and a smaller and nearer one to the south by Miss Caroline Herschel in 1783. Messier, observing Le Gentil's nebula in 1764, remarked that its position and form had remained constant since its discovery, but observations in recent years have raised some sus- picion of change. In September 1847, Bond detected two dark rifts running nearly parallel to the longer axis of the great nebula. For many years the significance of these channels remained a matter. of mystery, but a photo- graph taken by Dr. Isaac Roberts in December 1888 revealed at last their true character. They are now seen to represent the dark intervals between con- centric nebulous rings into which the nebula is divided. This wonderful photograph which will THE GREAT NEBULA IN ANDROMEDA 31 MESSIER. From the Original Negative taken by DR. ROBERTSON, Dec. 29^, i8i GREAT NEBULA. 199 mark an epoch in astronomical research shows us this great nebula for the first time in a clearly "intelligible form," and calls to mind the Nebular Theory of Laplace, in which the planets of the Solar System are supposed to have been evolved from rings detached from a rotating nebulous mass. In Dr. Roberts' photograph the nebula appears as a lengthened ellipse with a bright central nucleus. Its figure suggests that its real form is that of a circular mass, or rather a disc of large diameter but com- paratively small thickness like Saturn's rings projected into an ellipse by the high inclination of its plane to the background of the heavens. The dimensions of the nebula, as shown with cer- tainty on the photograph, are i 51' in length and about 23^-' in width. From these measure- ments I find that its apparent area is nearly three times greater than that of the full Moon. Such an extension would not, of course, be traceable in ordinary telescopes, but with a 1 5-inch refractor, Bond saw it still further extended. Assuming, how- ever, the more reliable measurements indicated by the photograph, we may make an attempt to ascer- tain the probable size of this wonderful object. To do this with accuracy, it would, of course, be necessary to know the exact distance of the nebula from the Earth. But, unfortunately, such knowledge is not yet available. The nebula has not hitherto afforded any evidence of proximity to our system. The new star 200 THE WORLDS OF SPACE. which suddenly blazed out near the nucleus in August 1885 was carefully measured for this pur- pose, but refused to reveal the secret of its distance. As the evidence of the spectroscope tended to show that the star was in the nebula, we may conclude that the distance is so great as to be practically immeasurable by our most refined and delicate methods of observation. Assuming, however, a minimum distance, such that light would take 160 years to traverse (corresponding to a parallax of one- fiftieth of a second of arc), we can easily calculate the dimensions of the nebula. At the great distance assumed, the apparent length shown by the photo- graph would indicate an actual diameter of no less than 333,000 times the Sun's distance from the Earth ! Light would therefore take over five years to pass from one side to the other of this vast nebula ! It has been suggested that the Andromeda nebula may possibly represent an external universe, but the improbability of this hypothesis will appear from the consideration that even the great diameter above computed is not much greater than the distance from the Sun to the nearest fixed star a Centaurf, and the limits of our universe are certainly vastly further from us than this. In fact, the visible universe, of which our Sun forms a member, most probably ex- tends far beyond the distance I have assumed for the Andromeda nebula. To suit the hypothesis of an external universe, we should, therefore, be obliged to GREAT NEBUUE. 2OT increase the distance of the nebula enormously, and this would lead us to an extravagant estimate of the real size and brightness of the new star above referred to. It seems, therefore, very improbable that this object forms an external galaxy. It is more prob- ably a member of the vast sidereal system in which our solar system is situated ; a system which, in all likelihood, includes the whole of the stars and nebulae visible in our largest telescopes. On the assumption that the nebula is a circular disc seen obliquely, we can also determine its posi- tion in space. Making the necessary calculations, I find that the plane of the nebula is inclined at an angle of 78 to the background of the sky, and that it probably lies at right angles, or nearly so, to the general plane of the Milky Way. The actual constitution of this marvellous object still remains a matter of mystery. The highest powers of the largest telescopes have hitherto failed to resolve it into stars. Yet the spectroscope shows that it is probably not gaseous, the spectrum being continuous, like that of the great globular cluster in Hercules. The component stars may possibly be comparatively small bodies, too small to be individ- ually visible even with our largest telescopes. Placed at the distance I have assumed for the Andromeda nebula, I find that our Sun would be reduced in brightness to a star of 9^ magnitude. If we assume the components to have only one-hundredth of the" 202 THE WORLDS OF SPACE. Sun's diameter (8660 miles), they would shine as stars of only 19! magnitude, which no telescope yet con- structed would show as individual points of light. Of much more irregular form than the Andromeda nebula is the great nebula in the " sword " of Orion. Its cloud-like appearance might perhaps suggest a different physical constitution, and the spectroscope shows it to be a mass of glowing gas. It has been called the " Fish-mouth " nebula, from the fantastic form of its central portion. It is just visible to the unaided eye on a very clear, moonless night as a glow round the central star of the " sword," and even with an opera-glass it is a conspicuous object. It seems curious, therefore, that it should have escaped the inquiring eye of Galileo, who paid especial attention to Orion, and that its existence should have remained unknown till the year 1618, when it seems to have been first seen by Cysatus, a Swiss astronomer. A drawing of it was published by Huygens in 1659, and it has been carefully studied and mapped by several observers in recent years. Sir John Herschel, during his visit to the Cape of Good Hope in the years 1834 1838, carefully observed and sketched the nebula, and his drawing published in the Cape Observations is an elaborate and valuable one. He says "The brightest portion offers a resemblance to the head and yawning jaws of some monstrous animal, with a sort of proboscis running out from the snout." Excellent drawings Exposure 81 minutes. Exposure 3-j hours. THE GREAT NEBULA IN ORION. GREAT NEBULA. 203 were also made by Bond in America, and Lassell at Malta. A comparison between the earlier drawings and those of later date has suggested the idea that great changes of form have taken place in the nebula; but probably the observed discrepancies may be simply due to different methods of delineation, rather than to changes which would, indeed, be necessarily on a gigantic scale to be perceptible at all in the comparatively short period during which the nebula has been accurately observed. Photography will, however, doubtless decide ere long whether such rapid changes are really in progress. The nebula has been very successfully photo- graphed by Dr. Common and Dr. Roberts, and these photographs confirm the general accuracy of the later drawings, but show a greater amount of detail. A remarkable nebula surrounding C(the southern star of the belt) has also been photographed by Mr. W. H. Pickering and Dr. Max Wolf. Mr. Pickering says "These plates show that it (the nebula in Orion) not only includes the sword handle c, i and 0, but a long nebulosity extends south from C, others surround this star, while others, both north and south, indicate that perhaps the next increase in sensitive- ness of our plates will join them all in a vast nebula many degrees in length." Near the densest part of the nebulous glow in the " sword " of Orion lies a remarkable multiple star, known as Orionis, the four brighter components of 2O4 THE WORLDS OF SPACE. which form the familiar " trapezium " of telescopists. I have seen these in the Punjab sky with a 3-inch telescope reduced in aperture to ij inch. Struve in 1826 detected a fifth star, and Sir John Herschel a sixth in 1830. Alvan Clark discovered a faint star within the " trapezium," and Mr. Barnard another with the great Lick telescope. Mr. Barnard has also found a faint double star just outside the trapezium, which Mr. Burnham finds a different object even with the giant telescope ! Indeed, the whole region in which the nebula lies is sprinkled over with faint stars. Nearly a thousand were observed by Bond in a portion covering about 3^. Observing the nebula with his great telescope, Lord Rosse thought that the nebulous light showed symptoms of incipient resolution into stars, but this view of its constitution has been completely over- thrown by the spectroscope, which shows it to consist of nothing but luminous gas. This was first demon- strated by Dr. Huggins. Referring to his earlier observations he says " The light from the brightest parts of the nebula near the trapezium was resolved by the prisms into three bright lines, in all respects similar to those of the gaseous nebulae. The whole of this great nebula, as far as lies within the power of my instrument, emits light which is identical in its character. The light from one part differs from the light of another in intensity alone." Further obser- vations fully confirmed this conclusion, and showed GREAT NEBUL.K 2O5 that one of the constituents of the nebulous matter is certainly hydrogen gas, of which the complete series of spectral lines has recently been photographed by Dr. Huggins. In his first observations, Dr. Hug- gins was disposed to identify the brightest line the " chief nebular line " as it is called with a line in the spectrum of nitrogen, but recent careful measures have shown him that the nebular line does not really coincide with the nitrogen line, but is distinct from it. Professor Lockyer considers that this line coincides with the edge of a fluting in the m.ignesium spectrum, but Dr. Huggins and Mr. Keeler find that such is not the case. Although undoubtedly very close to the magnesium line, it is separated from it by a small but distinct interval. Spectroscopic observations tend to show, in Dr. Huggins' opinion, that "the stars of the trapezium are not merely optically connected with the nebula, but are physically bound up with it, and are very probably condensed out of the gaseous matter of the nebula " another argument in favour of Laplace's Nebular Hypothesis, in which suns and planetary systems are supposed to have been evolved out of nebulous matter. In 1886, Dr. Copeland detected a line in the yellow portion of the spectrum, which he found to coincide with a line in the solar spectrum visible during total eclipses of the Sun. This solar line has not been identified with that of any terrestrial substance, and is supposed to indicate the presence in the Sun of some unknown substance, 2O6 THE WORLDS OF SPACE. to which the name " helium " has been given. Dr. Copeland says " The occurrence of this line in the spectrum of a nebula is of great interest, as affording another connecting link between gaseous nebulae and the Sun and stars with bright line spectra, especially with that remarkable class of stars of which the first examples were detected by M.M. Wolf and Rayet in the constellation of Cygnus." 1 Somewhat similar in its general appearance, and probably similar also in its chemical constitution, is the great nebula in the southern constellation Argo. This object, known as the "key-hole" nebula, sur- rounds the famous variable star rj Argus, a star which has fluctuated through all grades of brilliancy from that of Sirius to complete invisibility with the naked eye. This wonderful nebula lies in a very luminous portion of the Milky Way, and is thus described by Sir John Herschel : " It is not easy for language to convey a full impression of the beauty and sublimity of the spectacle which this nebula offers, as it enters the field of view of a telescope fixed in Right Ascen- sion, by the diurnal motion, ushered in as it is by so glorious and innumerable a procession of stars, to which it forms a sort of climax, and in a part of the heavens otherwise full of interest." 2 Nearly in the middle of the brightest part of the nebula lies the remarkable opening known as the " key-hole," or 1 Monthly Notices, R.A.S., June 1888. 2 Outlines of Astronomy, p. 653. THE T) ARGUS REGION AND NEIGHBOURING CLUSTERS IN THE SOUTHERN MILKY WAY. From a Photograph taken by MR. H. C. RUSSELL, Director of the Sydney Observatory, z^rd July, 1890. GREAT NEBULAE. 2O/ Icmniscate. The southern portion of this curious vacuity is completely free from nebulous light, as shown in Sir John Herschel's drawing, but the northern end is partly filled in with faint nebulosity. Several other somewhat similar vacuities are also visible. Although numerous small stars are scattered over the nebula, Herschel found no tendency to resolution in the nebula itself. He says " In no part of its extent does this nebula show any appearance of resolvability into stars, being in this respect analogous to the nebula of Orion. It has, therefore, nothing in common with the Milky Way, on the ground of which we see it projected, and may therefore be, and not improbably is, placed at an immeasurable distance behind that stratum." The evidence of the telescope has been confirmed by the spectroscope, which shows it to consist of luminous gas, like the Orion nebula. The position of some of the stars, however, with reference to the surrounding nebulous light suggests a real, and not merely an apparent, connection. The nebula with its branches covers an area of about one square degree, or about five times the area of the full moon. If placed at. the distance I have assumed for the nebula in Andromeda, it must fill a vast extent of space, a space compared with which our Solar System sinks into insignificance. A comparison of drawings has suggested considerable changes in the Argo nebula both in form and brightness ; but probably in the earlier delineations the fainter details were 208 THE WORLDS OF SPACE. obscured by the great brilliancy of ?? Argus when near its maximum. As the variable star is at present only of the seventh magnitude, the nebula is very visible to the naked eye, being, according to Mr. Abbott, brighter than the larger "Magellanic cloud." He says " In the twilight it appears as soon as a star of the second or third magnitude, the light being white and more diffuse, very like a small white woolly cloud in a blue sky, seen in sunlight." Much smaller than the Argo nebula, but somewhat similar in its general form, is another southern nebula known as 30 Doradus. This wonderful object forms one of the many varied forms which, collected together, compose the larger " Magellanic cloud." It is sometimes called the " looped nebula," from the curious convolutions formed by its nebulous rays and streams. Sir John Herschel gives a beautiful drawing of it in his Cape Observations, and describes it as " one of the most singular and extraordinary objects which the heavens present." Near its centre is a star of the ninth magnitude, attended by several fainter stars forming a small cluster. Just south of this is a pear- shaped vacuity, similar to the " key-hole " perforation in the Argo nebula, but of proportionately larger dimensions. North of the central star there is a bright nebulous ray like the tail of a comet. Near this is a round hole, and further north three plume- like branches diverge from a common nebulous stem. The? fainter portions of the nebulosity seem pierced GREAT NEBULA. by similar " coal sacks," features which appear to be characteristic of these large irregular nebulae. To satisfactorily explain the existence of these curious openings is a matter of no small difficulty. If we suppose the nebula to have a thickness compar- able with its visible extent, or, in other words, that it is of a roughly spherical form, we must suppose these vacuities to represent tunnels through a gaseous mass, a not very conceivable arrangement. If, however, we consider the nebula to have but little extension in the line of sight, that is, to form a thin stratum instead of a spherical mass, a perforation through such a disc is perhaps more easily imaginable. In either case, however, it is not easy to understand how an opening through a gaseous mass can be kept open, and prevented from closing up by fluid pressure. Many faint stars are scattered over the area occupied by 30 Doradus, and of these Sir John Herschel gives a catalogue of 105, ranging in brightness from the ninth to the seventeenth magnitude. I am not aware whether the light of this nebula has been examined with the spectroscope, but from its general appear- ance it will, I think, probably prove to be gaseous. Both the Argo nebula and 30 Doradus are un- fortunately invisible in these latitudes, but there are some other interesting objects further north. Of these may be mentioned that known as 8 Messier, which lies a little to the south-east of the fifth mag- nitude star 4 Sagittarii, a star close to the Ecliptic. 2IO THE WORLDS OF SPACE. I found it very plain to the naked eye in the Punjab sky, and even with a telescope of only 3 inches aperture it is a glorious object. The sixth magnitude star 9 Sagittarii is involved in the nebulosity which is partly mixed up, apparently at least, with a fine cluster of stars. I have seen the cluster well with a 3 -inch telescope. In the Cape Observations Sir John Herschel gives a drawing of this fine object, which shows several vacuities similar to those in the larger nebulae already described. He says "Its brighter portion may be described as consisting of three pretty distinct streaks or masses of nebula of a milky or irresolvable character, arched together at their northern extremities so as to form some resemblance to the arches of an italic letter m very obliquely written, and this is the aspect under which it strikes the eye on a cursory view. On closer attention these streaks are seen to be connected and run into each other below (or to the south), by branches and projections of fainter light, and to form three distinct basins, insu- lating oval spaces, one entirely, the other compara- tively, dark." He estimated the area covered by the nebula and its branches as about one-fifth of a square degree, or about the apparent area of the full moon. Secchi found a gaseous spectrum. About a degree north of the nebula just described lies a very curious object known as the " trifid nebula." It consists of three lobes or masses of nebulosity, separated by three dark lanes or rifts radiating from GREAT NEBULA. 211 the centre near which is situated a triple star. A beautiful drawing of this nebula has been made by Trouvelot, in which he shows one of the dark rifts curving round towards the north and separating the principal portion of the nebula from another nebulous mass which contains a central star. Sir John Herschel's drawing agrees fairly well with Trouvelot's, but does not show quite so much detail. The nebula has certainly a very nebulous appearance, but the spectrum is said to be not gaseous ! Sir John Herschel suspected a connection between the " trifid nebula " and 8 Messier. This region has been recently photo- graphed by Mr. Barnard, of the Lick Observatory, and by Mr. Russell, at Sydney, New South Wales. All the nebulae described in the preceding pages lie in or close to the Milky Way an interesting fact which seems to suggest that they form part and parcel of that wonderful zone. The nebulae in general show little or no signs of "proper motion," that is, motion across the face of the sky, a fact which suggests a vast distance from the Earth. In the case of the Orion nebula, Mr. Keeler, of the Lick Obser- vatory, has recently detected with the spectroscope a motion of recession in the line of sight at the rate of about 107 miles a second, but this is probably due, in part at least, to the Sun's known motion through space in the opposite direction. XXI. SPIRAL NEBULAE. THE Spiral Nebulae are among the most marvellous and mysterious objects in the heavens. They were discovered by the late Earl of Rosse, with his giant 6-feet reflector, and their spiral character has been fully confirmed by Dr. Roberts' photographs. Indeed photography has recently shown some nebulae as spirals which had not been recognized with the tele- scope as belonging to this interesting class. Perhaps the most remarkable of these wonderful objects is that known as 5 1 Messier, which is situated in the constellation Canes Venatici, about 3 south- west of t] Ursae Majoris, the star at the end of the Great Bear's tail (or handle of "the Plough"). It may been seen in small telescopes, but these merely show two nebulae of unequal size nearly in contact. Admiral Smyth, in his Celestial Cycle, describes it as " a pair of lucid white nebulae, each with an apparent nucleus, with their nebulosities running into each other, as if under the influence of a condensing power." SPIRAL NEBULA. 213 Sir John Herschel saw the larger of the pair as a bright globular mass, surrounded by a nebulous ring, the ring being partially split. He thought it might possibly be an external universe, similar to our visible universe, the luminous halo corresponding to the Milky Way in our stellar system. This hypothesis was not, however, verified by large telescopes, which show the object to-be really a complicated spiral. The nebula has been carefully drawn by Lord Rosse, Lassell, and Vogel, and their drawings have been substantially confirmed by photography. A photograph taken by Dr. Roberts on April 21, 1889, with an exposure of four hours, shows a bright central mass, encircled by spiral streams ; one of these streams being traceable from the centre to the smaller of the two nebulae, which looks as if it were on the eve of being thrown off from the parent mass by the force of the rotation. On the original negative, a tendency to a spiral structure is also seen in this small nebulae. Several still smaller nebulous masses or stars are visible in the field of view, and along the course of the spiral streams. This nebula was also photographed some few years since by Dr. Common at Ealing. Dr. Huggins finds that the spectrum is continuous. It would therefore seem to be not gaseous, and on the Nebular Hypothesis it would evidently represent a solar system in an advanced stage of its formation. The nebula known as 99 Messier in Virgo was also found by Lord Rosse to have a distinct spiral form 214 THE WORLDS OF SPACE. His drawing reminds one of a " Catherine wheel " in pyrotechnic displays. Key found it resolvable with an 1 8-inch reflecting telescope, and this observation was confirmed by D'Arrest. A photograph by M. Von Gothard shows the spiral branches very distinctly, and seems to indicate that the plane of the spiral lies nearly at right angles to the line of sight. Another spiral nebula lies a little south of the star A Leonis. Lord Rosse thought it resolvable into minute stars. A photograph of the nebula known as 81 Messier in Ursa Major, taken by Dr. Roberts in March 1889, with an exposure of four hours, shows a bright central mass, with a brighter nucleus, surrounded by spiral streams. In a paper print from the original negative, kindly sent to me by Dr. Roberts, these streams are very faint, but clearly traceable. It is evident from the photograph that the plane of the spiral is inclined to the line of sight, like the orbit of a binary star. As in the case of 51 Messier, small nuclei, or stars, are traceable along the spiral branches, but there is no outstanding small nebula ; and on the whole it is of a similar and more regular outline than the nebula in Canes Venatici. It may possibly be a solar system in an earlier sLage of its development. The spiral character of this object was unknown to astronomers before it was photographed. Dr. Huggins finds the spectrum continuous (or not gaseous), but, like the great nebula in Andromeda, only faintly visible at the SPIRAL NEBULAE. 215 red end. Like the Andromeda nebula also, it is not resolvable into stars with any telescope. Closely north of the above lies another curious object which Dr. Roberts believes to be a spiral nebula seen edgeways. It was described by Sir John Herschel as a " beautiful ray," and Lord Rosse speaks of it as " a most extraordinary object, at least 10' in length, and crossed by several dark bends." Ingall in 1885 described it as resembling " a distaff of flax." Nuclei and dark channels are perceptible in the photograph, and its general appearance is very much like Ingall's description. Dr. Huggins finds the same spectrum as in 81 Messier, so that we may, with much probability, conclude that the two nebulae are similar in character, and possibly differing only in their position in space. The great nebula in Andromeda (31 Messier), " the queen of the nebulae," as it has been aptly called, is visible to the naked eye, and was possibly known to the ancients. It is referred to by Al-Sufi, a Persian astronomer of the tenth century, as a familiar object in his day. Although the spectroscope shows that it is not gaseous, the largest telescopes have hitherto failed to resolve it into stars. It is of an elongated elliptical shape, and of considerable apparent size. The American astronomer, G. P. Bond, traced it to a length of 4, and a breadth of 2|, and observed two dark rifts running nearly parallel to the longer axis of the oval. In a beautiful photograph of this 2l6 THE WORLDS OF SPACE. object recently taken by Dr. Roberts at Liverpool, Bond's rifts are seen to be really dark spaces between the central nucleus and surrounding rings. Here we seem to see a system in a more advanced stage of its development than appears to be the case in 5 1 Messier. The spiral character is nearly lost, and the spiral branches having become detached from the parent mass, have now formed rings encircling the nucleus. Dr. Roberts says " they present a general resemblance to the rings of Saturn/' and that the nebula " is now for the first time seen in an intelligible form." Two adjoining nebulae are visible on the photograph one completely separated from the great nebula, and the other nearly so, and we are tempted to imagine with Dr. Roberts that these represent portions of the large nebula which have been detached, and are now " undergoing their transformation into planets." The truth of Laplace's Nebular Theory seems to be here clearly demonstrated. The accompanying plates are photographs of spiral nebulae taken by Dr. Roberts in April 1893. They have been reproduced from glass positives kindly sent to me by Dr. Roberts. In the Monthly Notices of the Royal Astronomical Society for December 1893, Dr. Roberts describes these photographs as follows : y I 1 68 Ursce Majoris. "The photograph shows this to be a very interesting spiral nebula, almost perfect in outline. In the centre of the spiral is a star I 205 URS^E MAJORIS g* i 4 m + 31* 34'. April izth, 1893. 3^ hours exposure. I 168 URS^E MAJORIS io'i 12 + 41" 57'. SPIRAL NEBUL.K 217 of fourteenth to fifteenth magnitude, and around it are formed the convolutions, each of which is broken up into stars ; four of them (omitting the bright star on the north side) are sharply defined, and the others, which are numerous, appear to be in all stages of development, between very faint star-like patches and the defined ordinary star images." "There is still nebulosity between some of the spirals, as well as between the stars in the con- volutions." " Several photographs of spiral nebulae have from time to time been presented to the Society, and each one of them shows the spirals to be broken up into stars, or into star-like condensations, and I think the cumulative evidence thus brought before us amounts to a demonstration of the formation of stars by the condensation of nebulosity, or by the aggregation of meteoric or other cosmic matter." y I 205 Ursa Majoris. " The enlarged photo- graph, and more clearly the negative, show the nebula to be a symmetrical ellipse, with a distinctly stellar nucleus in the midst of dense nebulosity which surrounds it. Outside this is a well-defined zone of faint nebulosity, and then a broad ring, or zone, with little if any nebulosity in it. Outside this, again, is a very dense broad ring of nebulosity, and a patch of very faint nebulosity extends beyond the ring at the j./end." " The nebula is probably a circular or oval system 218 THE WORLDS OF SPACE. seen in perspective elongated, and there are indications of condensations of matter of the ring." " The nebula resembles, on a very small scale, the great nebula in Andromeda, and the description given by Lord Rosse agrees well with the photograph, though he could not have seen the details which are there shown." The scale of the photographs is I millimetre to 24" of arc. XXII. PLANETARY NEBUL.-E. PLANETARY nebulae are among the most remark- able and interesting objects in the heavens. They were so named by Sir W. Herschel from their resem- blance to planetary discs. This supposed similitude applies only to their general uniformity of brightness, and to the absence of any definite nucleus or brighter portion, at least when viewed with telescopes of no great power. In intrinsic brilliancy they are of course vastly inferior to the planets, being usually very faint objects, and only to be well seen with powerful telescopes. About three-fourths of the known planet- ary nebulae are situated in the Southern Hemisphere J but there are some interesting examples north of the Equator. One of the most remarkable of these wonderful objects is that known as 97 Messier. It was dis- covered by Me"chain in 1781, and lies about 2 (four diameters of the Moon) to the south-east of the star Ursse Majoris the southern of the two 220 THE WORLDS OF SPACE. "pointers." It has an apparent diameter so large, that if we consider it placed at the distance of the nearest fixed star, it would fill a sphere the diameter of which would be about three and a half times that of the orbit of Neptune. In Sir W. HerschePs tele- scope its light appeared nearly umform ; but Lord Rosse's great reflector has disclosed the existence of two openings with apparently a stellar point in each vacuity. From this peculiarity it has been termed the " Owl nebula." One of the stars seems to have disappeared since the year 1850, a fact which adds to its interest and mystery. Dr. Huggins finds the spectrum gaseous. Close to the pole of the ecliptic (between 8 and Draconis) is a remarkable planetary nebula, known to astronomers as Herschel iv. 37. Smyth describes it as "a remarkably bright and pale blue object." Webb saw it as " a very luminous disc, much like a considerable star out of focus." Sir W. Herschel saw a very small nucleus which Bird estimated as equal to a tenth magnitude star in 1863, and D' Arrest to one of the eleventh. Examined with the great Lick telescope, Professor Holden found it "apparently composed of rings overlying each other," which he thinks are probably " arranged in space in the form of a true helix." The diameter of the nebula is about 20". Brunnow found a parallax of O'O47". This would make its actual diameter about 425 times the Sun's distance PLANETARY NEBULA. 221 from the Earth, or about 39,478,000,000 miles ! Bredechin, however, finds a parallax of only O'oog", which would indicate a diameter of 2222 times the Sun's distance, or mi the diameter of the Earth's orbit ! Such are the stupendous objects which astro- nomy reveals to our wondering gaze. A little to the south-west of the fifth magnitude star 5 1 Herculis (between /3 and 8) is a small planetary nebula, discovered by Struve. D'Arrest estimated its light as equal to that of a star of the eighth magni- tude. There are several stars in the field with which the light of the pale blue nebula can be conveniently compared. Webb found it very bright, but not sharply defined, and as Herschel described it exactly like a star out of focus. In Lord Rosse's telescope it appeared of an intense blue. Secchi thought he could resolve it into minute stars with a magnifying power of 1500; but in this he was evidently mistaken, for the spectroscope has decided that its nature is gaseous. Another interesting object of this class is found closely preceding the fifth magnitude star v Aquarii. It was discovered by Sir W. Herschel in 1782, and forms one of nine rare celestial objects given by Struve, who saw it of a well-marked elliptical form. Admiral Smyth described it in 1836 as "bright to the very disc, and but for its pale blue tint, would be a very miniature of Venus." Within the larger nebu- losity Lassell could detect a brighter elliptic ring. 222 THE WORLDS OF SPACE. Secchi suspected it to be a mass of small stars ; but Dr. Huggins' spectroscope shows it to be nothing but luminous gas. It has an apparent diameter of about 20" of arc, and if placed at only the distance of a Centauri it would more than fill the orbit of Saturn. Examined with the great telescope of the Lick Observatory, Professor Holden found it a wonderful object of a pale blue colour, with a central ring, which he compares to a "footprint left in the wet sand on a sea-beach." About 4j following y Eridani lies another re- markable planetary nebula. Smyth describes it as " a splendid though not very conspicuous object of a greyish-white colour. It is somewhat like a large star out of focus with a planetary aspect." Sir W. Herschel found it somewhat elliptical and not well defined, and thought it might possibly be a very compressed cluster of stars at an immense distance from the Earth. Observing it at the Cape of Good Hope, Sir J. Herschel concluded that it is " a very distant and highly-compressed cluster." These views have been confirmed with the spectroscope by Dr. Huggins, who found that the spectrum, although deficient at the red end, like the great nebula in Andromeda, is not gaseous. Lassell describes it as the most wonderful object of the kind he had ever seen an apparently stellar point in the centre of a circular opening which is again surrounded by a fainter nebulosity. D'Arrest thought the edges re- PLANETARY NEBULA. 223 solvable, and Lord Rosse considered the nucleus to be granular in his great reflector. About 2 south of the star /x Hydrae is a planet- ary nebula which Smyth describes as resembling Jupiter in "size, equable light, and colour." He thought that, " whatever be its nature," it " must be of awfully enormous magnitude." Webb found it of " a steady pale blue light," and Sir J. Herschel at the Cape found its colour "a decided blue" a feature which seems very characteristic of these extra- ordinary objects. Sir W. Herschel failed to resolve it into stars ; Secchi found it broken up into appar- ently stellar points and a sparkling ring ; but Dr. Huggins finds the spectrum gaseous, so that Secchi's points can scarcely be stellar. Many examples of this interesting class of nebulae were observed by Sir John Herschel during his resid- ence at the Cape of Good Hope. His descriptions of some of these are given in my Scenery of the Heavens }* The gaseous character of most of these planetary nebulae led Professor Pickering to the conclusion that possibly an examination with the spectroscope might lead to the discovery of objects of this class, which owing to their small apparent diameters might escape detection with the telescope, and merely appear like small stars, even when examined with telescopes of considerable power. The light of an ordinary star is seen spread out by the prism into a 1 pp. 210 212. 224 THE WORLDS OF SPACE. coloured band of light ; but the light of a gaseous nebula, being chiefly of one colour, forms merely a small point or disc of light. Following this plan, a systematic search for such objects has been instituted at the Harvard Observatory (U.S.A.), a direct-vision prism being placed for the purpose between the eye- piece and the object-glass of the 1 5-inch refractor of that Observatory. A large number of stars have been carefully examined in this way, and the search has disclosed the existence of some small planetary nebulae hitherto unsuspected. In the first sweep* made on July 13, 1880, an object entered the field of view of the observing telescope, having a spectrum consisting of a bright point of light, quite unlike that of the ordinary stellar spectrum. This proved to be a small planetary nebula. It lies in the Milky Way near the star A. Sagittarii, and measures of its light showed that it was about the eleventh magnitude. Its disc is so small that probably its nebulous nature would never have been detected with the telescope. On the following evening a similar object was found in the same vicinity. It was some- what fainter than the first, but with rather a larger disc. This region of the heavens was selected for observation because it contains four out of the 50 planetary nebulae previously known to astronomers. On August 28 an object was met with, having a remarkable spectrum with " two bright bands near the ends of a faint continuous spectrum." PLANETARY NEBULA. 225 It was soon identified with the star Oeltzen Arge- lander, No. 17,681, which lies near ju Sagittarii. It was observed once by Argelander, and twice at Washington, and on each occasion was recorded as an ordinary fixed star. These observations show that it existed in its present position and brightness over thirty years previously. An examination of the remarkable spectrum shown by this object, indicates that the brighter parts are bands and not lines. One of these bands lies near the sodium line D of the solar spectrum, and another in the blue between F and G. Professor Pickering says the spectrum of this object " is unlike that of any other source of light as far as is yet known." " On the other hand," he says, " it resembles a star in other respects, showing no disc, and having a much greater intrinsic brightness than other nebulae." Another new object, similar to the preceding, was detected on September 2, 1880, and in the same region of the heavens. It resembles a star of the thirteenth magnitude, and is probably the faintest planetary nebula hitherto discovered. XXIII. CELESTIAL PHOTOGRAPHY. THE first idea of photographing the heavenly bodies may be dated back to the great discovery made by Daguerre and Niepce, and communicated by Arago to the French Academy on August 19, 1839. Daguerre attempted to photograph the Moon in 1840, but apparently without much success, and to Professor J. Draper seems due the credit of obtaining, in the same year, the first lunar photo- graph. He was also the first to photograph the solar spectrum. This was accomplished in 1843. The earlier photographs were called daguerreotypes, after the inventor, and a series of plates of an eclipse of the Sun, now almost obliterated by time, are preserved in the museum of the Paris Observatory. These plates have no date, but are probably photo- graphs of the total eclipse of July 8, 1842 an eclipse which was also photographed by Majochhi at Milan. In April 1845, Messrs. Fizeau and Foucault obtained a good photograph of the Sun with an CELESTIAL PHOTOGRAPHY. 227 exposure of one-sixtieth of a second, which showed some sun-spots, and also gave some indications of the fact, now well known, that the centre of the solar disc is brighter than the portions near the limb. The solar eclipse of July 28, 1851, was photo- graphed by Berkowski at Koenigsberg, and upon this plate are seen for the first time traces of the solar corona and red flames. In 1853 and 1854 Mr. Hartnup, at Liverpool, and others, obtained good photographs of the Moon, and the solar eclipse of May 26, 1854, was photographed by Professor Bartlett at West Point (U.S.A.). In 1856 the late Mr. Warren de la Rue established an observatory at Cranford, having a reflecting telescope made by himself of 13 inches aperture and 10 feet focal length. In the following year he obtained with this instrument good photographs of the Moon, Jupiter, and Saturn. It may be mentioned here that by photographing the Moon at suitable intervals, and taking advantage of the small changes caused by libration, it is possible to obtain pictures which, when viewed in a stereoscope, give the appearance presented by a spherical body. In 1859 Mr. de la Rue constructed a large photo- heliograph at Kew, and with this instrument over 2000 photographs of the Sun's surface were obtained in the years 1862 to 1872. In the years 1850 to 1871 six eclipses of the Sun were photographed with varying success, but most of them showed details of 228 THE WORLDS OF SPACE. the corona and protuberances. Photography was also applied to the observation of the transits of Venus in 1874 and 1882, but the results obtained in this way were of doubtful value. The first attempt to photograph the stars seems to have been made in 1850 by W. C. Bond and Whipple, with the equatorial telescope of the Harvard Observatory (U.S.A.). They succeeded in obtaining images of the bright star Vega and the double star Castor, but the daguerreotype plates used were not sufficiently sensitive for such delicate work. A very long exposure was necessary for even a bright star, and the impossibility of obtaining impressions of the fainter stars led them to abandon their efforts. On the introduction of the collodion process, however, in 1857, Bond again attempted to photograph the brighter stars, and succeeded in obtaining images of stars of the first and second magnitude with an exposure of only two seconds. He also photo- graphed the double star Mizar (f Ursse Majoris), and from measures of the photographed images he ob- tained a very accurate result of the position, angle, and distance between the components of this well- known pair, a result which agreed closely with Struve's measures made directly with the micrometer. In the year 1864 Dr. Rutherford, of New York, constructed an object glass of II inches aperture for the photography of the stars, which gave better results than a mirror of over 23 inches in diameter. CELESTIAL PHOTOGRAPHY. 229 With this instrument he obtained good photo- graphs of the Praesepe and the Pleiades, which showed stars down to the ninth magnitude. Follow- ing Rutherford's methods, Dr. Gould, in the years 1870 to 1882, photographed the principal star clusters in the Southern Hemisphere at the Cordoba Observ- atory in the Argentine Republic, and obtained plates showing stars to the tenth and eleventh magnitudes, some of them containing 500 stars to the square degree. We now arrive at an epoch when the photography of celestial bodies was much facilitated by the intro- duction of the dry plate process. The old wet plate process with collodion required enormously long ex- posures, and was difficult to manipulate. The dis- covery of the gelatino-bromide of silver process, however, allowed much more sensitive plates to be used, thus reducing the time of exposure necessary for faint stars, and permitting more work to be done in a given time. The advantage of a short exposure in a climate like that of England is obvious, for with a long exposure the sky may cloud over before the photograph is finished. The introduction of the dry plate process gave a fresh impetus to celestial photo- graphy. In the year 1881 Professor Draper obtained a good photograph of the great nebula in Orion with an exposure of one hour, and a still better photo- graph of this wonderful object was obtained by Dr. Common in January 1883. 230 THE WORLDS OF SPACE. About this time Professor Pickering, at the Harvard Observatory, commenced to take a series of plates showing all stars to the naked eye, and this work is now far advanced towards completion. The idea of forming a complete chart of the whole sky by means of photography seems to have been originally suggested by Mr. de la Rue about the year 1857, but the possibility of carrying out the idea does not seem to have been recognized until Dr. Gill, in 1882, obtained a photograph of the comet of that year with an exposure of I hour 50 minutes. On this photograph numerous faint stars are sharply shown, and Dr. Gill's success proved that a photographic chart of the sky was now possible. The same sug- gestion seems to have been also made by Dr. Common in the same year. In the year 1884, Messrs. Paul and Prosper Henry, of the Paris Observatory, while attempting to complete by eye observations at the telescope the ecliptic charts left unfinished by Chacornac at his death in 1873, came across regions of the Milky Way so rich in the fainter stars, which it was necessary to show on these charts, that they found it impossible to record them in their correct places. The idea then occurred to them that the work might be successfully accomplished by the aid of photography. To carry out this plan they constructed a special photographic equatorial telescope with an object glass of 13 inches aperture, and with lenses corrected for photographic work. To this in- CELESTIAL PHOTOGRAPHY. 231 strument is attached a second telescope of about 9-*- inches aperture, which serves as a finder in following the stars to be photographed, and corrects any error in the rate of the clock which drives the whole appar- atus. Complete success crowned the efforts of these admirable astronomers, and the excellence of the stellar photographs obtained can only be appreciated by those who have seen the originals. The present writer has in his possession a number of paper prints from the original negatives, very kindly sent to him by the brothers Henry. On these the stars are shown as perfectly round white discs on a black background, and their extreme beauty can hardly be adequately described. The duration of exposure necessary for obtaining images of stars of different magnitudes varies, of course, with the brightness. At the Paris Observatory it has been found that, with the extra sensitive plates now used for the purpose, stars of the first magnitude can be photographed with an exposure of -g-J-oth of a second, and stars of the second magnitude in about T frth of a second. Stars of the sixth magnitude about the faintest visible to ordinary eyesight without optical aid take about half a second ; those of the ninth magnitude about eight seconds ; those of the twelfth about two minutes ; and for stars of the six- teenth magnitude about the faintest visible in the largest telescopes an exposure of i hour 20 minutes is required. The long exposures necessary to obtain 232 THE WORLDS OF SPACE. images of the fainter stars cause those of the brighter stars to be much " over exposed," and these latter appear on the photographs as discs of considerable size. There is, however, a relation between the size of the disc and the brightness of the star, and estima- tions of relative brightness can be obtained by measur- ing the diameter of the discs with an instrument which has been invented for the purpose. Stars of a red colour are, of course, more difficult to photograph, and the images of these ruddy stars come out with much smaller dies than those of white stars of equal brilliancy to the eye. The superiority of the photographic method of stellar mapping over eye observations will be under- stood from the fact that a chart of the Pleiades constructed by M. Wolf with a large telescope shows 67! stars in this well-known cluster, while on a photo- graph taken at the Paris Observatory with an exposure of three hours, over 2000 stars can be counted ! Indeed faint stars are visible on these stellar photographs which cannot be seen by the eye in the telescope with which the photographs were taken. In fact, the success attained at the Paris Observatory in stellar photography is such that, the late superintendent, Admiral Mouchez, remarked that "at first sight a doubt might be entertained as to the accuracy of the results." Good photographs have also been obtained by Dr. Roberts, using a reflecting telescope of 20 inches CELESTIAL PHOTOGRAPHY. 233 aperture. On a plate of a portion of the Milky Way in Cygnus, taken in August 1887, with an exposure of an hour, no less than 16,000 stars can be counted on a space of about four square degrees. Photographs of double stars have also been taken at the Paris Observatory, and from these accurate measures of position can be obtained which will be of great value in the calculations of the orbits of these interesting systems. To avoid the danger of accidental specks on the plates being mistaken for stars, the plan has been tried at the Paris Observatory of giving three ex- posures on each plate, the three images forming a small equilateral triangle, which can only be seen with the aid of a microscope. Another advantage of this method is that the time of exposure for each image may be shortened, the triple image producing a combination which is more easily seen than a single image, Should one of the minor planets between Mars and Jupiter happen to be present in the region photographed, it could at once be detected by the distorted shape of the small triangle ; and it is be- lieved that even a planet beyond Neptune might be detected in this way. If a trans-Neptunian planet exists ft will probably be discovered by photography. Several of the minor planets have already been dis- covered by this method. For the fainter stars, however, an exposure of one hour is necessary for each image, and the continuous 234 THE WORLDS OF SPACE. exposure of three hours required for the triple image is a great strain on the observer's eye and patience, and forms a serious objection to the general adoption of this plan. By the use of more sensitive plates and larger telescopes, the time of exposure will probably be considerably reduced in the future. When exam- ined with a microscope, however, the appearance of the stellar images is so peculiar that accidental spots can be easily distinguished. With a high power the images look like globular clusters of stars as seen in a telescope, and they have been described as resem- bling " a heap of shingle." For this reason a single exposure will probably be adopted in constructing a chart of the heavens. In the year 1887 an International Conference of astronomers was held at Paris, for the purpose of arranging the details of a photographic chart of the whole heavens on a uniform plan. The Conference was composed of fifty-six astronomers, representing sixteen different nations. It was resolved that re- fracting telescopes of about 13 inches aperture should be used in the work. A series of plates will first be taken showing stars to the eleventh magnitude. From these a catalogue will be formed, which will serve as a foundation and reference for all similar work in the future. This catalogue will, it is supposed, include about one and a half million of stars. A second set of plates will also be obtained showing stars to the fourteenth magnitude inclusive, CELESTIAL PHOTOGRAPHY. 235 and these will probably contain about 20,000,000 stars. Another Congress was held in Paris in September 1889, when further details were considered and arranged. The work will be divided among a number of ob- servatories in different parts of the world, and already some progress has been made in the construction of the necessary telescopes and the carrying out of the plan, which, it is hoped, will be brought to a success- ful conclusion before many years have elapsed. Some remarkable results have already been ob- tained by photography. The satellites of Mars and Neptune and Saturn's rings have been photographed. Photographs of the great Orion nebula by Drs. Common and Roberts show a great extension of this wonderful object. The great nebula in Andromeda, the spira-1 nebulae, and others, have been successfully photographed by Dr. Roberts. In the Andromeda nebula, we see that the whole mass is breaking up into rings of nebulous matter, and the photograph sug- gests unmistakably the probable formation of a solar system on the lines of Laplace's Nebular Hypothesis. Photography has also been successfully employed in the determination of stellar distances and in the study of stellar spectra. A catalogue of the spectra of over 10,000 stars, and entitled The Draper Cata- logue, has recently issued from the Harvard Observ- atory. It has been formed in honour of the late THE UNIVERSITY 1 236 THE WORLDS OF SPACE. Dr. Henry Draper, and forms a fitting monument to the memory of that distinguished astronomer. An examination of the photographed spectra has revealed the existence of some very close binary stars revolv- ing in unusually short periods double stars of which the components are so close that the largest telescopes in existence have failed to show them as anything but single stars. Photographs of the Milky Way have also been successfully taken by means of a portrait lens strapped to a telescope driven by clockwork. The pictures obtained are very beautiful, and give promise of still better results in the future. The study of these and the stellar charts, when completed, will probably add considerably to our knowledge of the construction of the sidereal universe. The preceding short sketch of the rise and progress of celestial photography will show that we may look forward in the future to a great increase of astrono- mical knowledge by its aid. Assisted by even larger telescopes and more sensitive plates than we now possess, the heavens will be studied in a way undreamt of by our ancestors, and, in the words of a well-known astronomer, " the records that future astronomers will use will not be the written impressions of dead men's views, but veritable images of the different objects of the heavens recorded by themselves as they existed." XXIV. ASTRONOMY WITHOUT A TELESCOPE. SOME years since Sir Edmund Beckett (now Baron Grimthorpe) published a book entitled Astronomy ^without Mathematics. Many may think that these two sciences could not be separated ; but it is quite true that much may be learned about the heavenly bodies without any deep knowledge of mathematical science. Another work has recently been published by an American writer Mr. Garrett P. Serviss with the title Astronomy with an Opera-glass, in which he shows how much may be seen with such modest " means " as an opera-glass of little over one inch and a half in diameter. With such an instrument, or one slightly larger, really good work may be done in the observations of the brighter variable stars, and it may be even used for the discovery of new objects of this class, which are not much below the range of naked- eye vision. In the following pages I propose to go a step further than Mr. Serviss has done, and to point out the knowledge of astronomy which may be acquired- 238 THE WORLDS OF SPACE. without an instrument of any kind, the observer being aided only by the naked or " unarmed " eye, or, as one of my American correspondents phrases it, the " undraped optic " ! I hope to show that something, at least, may be learned in this way of what Mr. Serviss terms " the celestial city, whose temples are suns, and whose streets are the pathways of light." It must be remembered that astronomy was studied ages before the invention of the telescope, and that the ancient astronomers gained, without any optical assistance, a considerable amount of knowledge respecting the heavenly bodies. Let us first consider the stars visible to the naked eye. The number of these down to the sixth magni- tude about the faintest that average eyesight can see is, for both hemispheres, about 6000. The number, therefore, visible at one time from any given place is about 3000. Possibly double this number might be seen by those gifted with exceptionally keen eyesight ; but even this is a comparatively small number, scattered as it is over so large an area. Those who do not possess the power of effective enumeration estimate the number visible to the naked eye as con- siderably greater than is really the case. This is partly due to the irregular distribution of the lucid stars over the celestial vault, and partly to the effect which the aspect of the starry sky produces on the imagination ; the fact of the stars increasing in number as they diminish in brightness inducing us to suspect ASTRONOMY WITHOUT A TELESCOPE. 239 the presence of points of light which we do not actually see. An attempt to count those visible with certainty in any selected portion of the sky will, how- ever, convince any intelligent person that the number, far from being large, is really very small, and that the idea, which some entertain, of a countless multitude, is merely an optical illusion, and a popular fallacy which has no foundation in fact. Of course the number visible in telescopes is very considerable. Perhaps with the largest telescopes 100,000,000 could be seen ; but even this large number is very far from being " countless." The present popula- tion of the earth is about 1,400,000,000, or about 14 times the number of the visible stars ! The first thing to be done in studying the heavens with the naked eye is to learn the positions and names of the brighter stars ; and from these the fainter ones may easily be identified by means of a star atlas. Those who study the stars in this way have probably a more intimate knowledge of the starry heavens than professional astronomers, who generally find the stars at least the fainter ones by referring to a catalogue of stars, and then setting their telescope to the place indicated by the figures given in the cata- logue. Although the famous astronomer Sir William Herschel possessed several large telescopes, he also studied the stars with the naked eye, and it is related of this great observer that he could without hesitation identify any star he could see in this way, by its 240 THE WORLDS OF SPACE. name, letter, or number ! Such an exhaustive know- ledge of the heavens is of course very rare ; but an acquaintance with all the brighter stars can easily be acquired by any person of ordinary intelligence. The " Plough," or Great Bear, is familiar to most people. This remarkable group of seven stars will be found very useful in identifying some of the brighter stars. The two stars furthest from the "tail" are called the " pointers," as they point nearly to the Pole Star, or star to which the axis of the earth nearly points. I say " nearly," for the Pole Star is not exactly at the pole, but distant from it about three diameters of the Moon. The northern of these stars is known to astronomers by the Greek letter Alpha, and the southern as Beta. The others, following the order of the figure, are known by the letters Gamma, Delta (the faintest of the seven), Epsilon, Zeta, and Eta. Now, if the curve formed by the three stars in the tail, Epsilon, Zeta, and Eta, is continued on, it will pass near a very bright star. This is Arcturus (Alpha of the constellation Bootes), one of the brightest stars visible in Europe. Again, if we draw an imaginary line from Gamma to Beta, and produce it, it will pass near another bright star. This is Capella (Alpha of Auriga, " the Charioteer " referred to by Tennyson). Again, if we draw a line from Delta to Beta, and produce it, it will pass near the tolerably bright stars, Castor and Pollux (Alpha and Beta of the constella- tion Gemini, or the Twins), the northern of the two ASTRONOMY WITHOUT A TELESCOPE. 241 being Castor. Another line from Delta to Gamma produced will pass near a bright star called Regulus (Alpha of Leo, the Lion). Another line from Beta to Eta will pass near a group called Corona Borealis, or the Northern Crown. On the opposite side of the Pole Star from the Plough, a group of five conspicuous stars will be found, forming a figure shaped somewhat like a W. This is Cassiopeia's Chair. Commencing with the most westerly of the five, these stars are known as Beta, Alpha, Gamma, Delta, and Eta. Like the stars of the Plough, those of Cassiopeia's Chair may be used to find other stars. For instance, a line drawn from Beta to Alpha passes close to a star known as Gamma in Andromeda ; and the same line produced in the opposite direction will pass a little north of the bright star Vega (Alpha Lyrae), one of the brightest stars in the northern heavens. A line from Gamma to Alpha produced will pass through the well-known " Square of Pegasus." To the east of Vega lies Cygnus, or the Swan, a well-known northern constellation. It may be recog- nized by the long cross formed by its principal stars, Alpha, Beta, Gamma, Delta, and Epsilon ; Alpha, or Deneb, being the most northern and brightest, and Beta the most southern and faintest of the five. To the south-east of Cassiopeia's Chair lies the constellation Perseus, distinguished by its well-known festoon or curve of stars. South of this lies the R 242 THE WORLDS OF SPACE. constellation Taurus, or the Bull, which contains the well-known groups or clusters, the Pleiades and the Hyades. The Pleiades form perhaps the most re- markable group of stars in the heavens, and are easily found, when above the horizon, in the winter months in England. To ordinary eyesight the cluster consists of six stars. Some persons gifted with exceptionally keen eyesight have, however, seen eleven or twelve. A map of the Pleiades made in the sixteenth century shows eleven stars very correctly. This was drawn, of course, from observations made with a measuring instrument, but without the aid of a telescope. The observer (I think it was Mostlin, Kepler's tutor) must have possessed wonderfully sharp eyesight. The Hyades form a V-shaped figure, and contain the bright reddish star Aldebaran. South of Taurus and Gemini will be found the splendid constellation of Orion, perhaps the most brilliant group of stars visible in either hemisphere. A remarkable quadrilateral figure is formed by its four stars, Betelgeuse (Alpha) and Gamma on the north, and Rigel (Beta) and Kappa on the south. Of these Betelgeuse and Rigel are bright stars of the first magnitude. Betelgeuse is distinctly reddish, and also slightly variable in its light. Rigel is a beautiful white star. In the middle of the quadri- lateral are three stars of the second magnitude, nearly in a straight line, known as Delta, Epsilon, and Zeta, Delta being the northern of the three. These form ASTRONOMY WITHOUT A TELESCOPE. 243 Orion's " belt." South of these are three fainter stars, also in a straight line, forming the " sword " of Orion. Surrounding the central star of the " sword " is " the great nebula in Orion," one of the finest objects in the heavens. It is barely visible to the naked eye, but may be seen with a good opera-glass. To the south-east of Orion will be found Sirius, the brightest star in the heavens. It is the chief star of the constellation Canis Major, or the Great Dog, and has been well termed "the monarch of the skies," from its great brilliancy. The bright star Regulus, referred to above, is situated in a remarkable group of stars shaped like a sickle, and known as " the Sickle in Leo." Regulus lies at the extremity of the handle. Leo is well placed for observation in April and May. With the help of the bright stars mentioned, and the aid of a star atlas, the other constellations may be easily identified. The famous group called the Southern Cross is not visible in England, but forms a conspicuous object in the southern heavens. It has formed a subject of interest since the earliest ages of antiquity. Its component stars are, however, not so brilliant as some suppose, the two brightest being between the first and second magnitude, the next of the second, and one between the third and fourth magnitudes. Near the Southern Cross are two bright stars known as Alpha and Beta of the Centaur. 244 THE WORLDS OF SPACE. Among the stars are many objects known as " double stars." These consist of two stars very close together, but which appear to the naked eye only as single stars. Some are triple, and even quadruple. Of these double stars there are now about 10,000 known to astronomers, but they are only visible with a telescope. Some, indeed, are so close that the highest powers of the very largest telescopes are necessary to see them as anything but single stars. Of the naked-eye stars there are, however, some apparently so close that they present very much the appearance of real double stars as seen in a telescope. These, although not recognized by astronomers as double stars, have been termed " naked-eye doubles." Houzeau found that the brighter the stars are, the easier it is to separate them ; and that for small stars, about 15' of arc, or half the Moon's apparent diameter, is about the limit below which the naked eye cannot see a faint star double. Of the " naked-eye doubles," perhaps the most remarkable is Mizar, the middle star in the " tail " of the Great Bear. Close to it is a small star, some- times called "Jack on the Middle Horse." It was known to the ancient astronomers as Alcor, or " the test," as it was then considered a test of excellent eyesight. Whether it has really brightened seems doubtful, but at present it is perhaps visible to ordinary eyesight. Some, however fail to see it, ASTRONOMY WITHOUT A TELESCOPE. 245 while to others with keener vision it seems as plain as the proverbial " pike-staff." The star Alpha Capricorni consists of two stars which, although closer than Mizar and Alcor, are more equal in brightness, and may be easily seen with the naked eye on a clear night. Nu Sagittarii may also be seen double in this way. Theta Tauri, in the Hyades, is another object which some eyes can see distinctly double ; also Kappa Tauri, a little north of the Hyades ; Omicron Cygni, a little to the west of Alpha Cygni (Deneb), is another example. On a very fine night two stars may be seen in Iota Orionis, the most southern star in the "sword." Near Gamma Leonis, one of the brightest stars in the " sickle," is a star of the sixth magnitude, which some can see without optical aid. The most severe test in the Northern Hemisphere is, however, Epsilon Lyrae, the northern of two small stars which form a little triangle with the brilliant Vega. This, to some eyes, appears double. The famous German astronomer Bessel is said to have so seen it at thirteen years of age. To most people, however, it will perhaps appear only elongated. This is a very remarkable star, as each of the components is seen to be a close double when examined with a good telescope, and between the pairs are several fainter stars. Among those interesting objects, the variable stars, are several which may be well observed without optical assistance. Of these may be mentioned 246 THE WORLDS OF SPACE. Algol, of which all the fluctuations of light may be easily observed with the naked eye ; Mira Ceti, which may be well observed when at its brightest ; Lambda Tauri, a variable star of the Algol type ; Betelgeuse (Alpha Orionis), which is slightly variable ; Zeta Geminorum, a fourth magnitude star, which varies about three-quarters of a magnitude in a period of about ten days ; R. Hydrae, which is visible to the naked eye at maximum ; Beta Lyrse, period about thirteen days ; Eta Aquilae, period about seven days ; and Delta Cephei, which varies about one magnitude in a period of a little over five days. Of all these stars useful observations may be made without optical assistance of any sort. Observations, and even discoveries, of new or "temporary" stars may also be made with the naked eye. This occurred in the case of the " temporary " stars of 1572, 1604, 1670, 1866, and 1876, but of course these were bright objects at the time of their discovery. Hind's "new star" of 1848 in Ophiuchus was, however, only of the fifth magnitude when it appeared, and it might have escaped detection with the naked eye. A star of this magnitude might, however, be easily detected by an observer who is familiar with the principal stars of a constellation. 1 The Milky Way may perhaps be better seen with 1 A new star was discovered near Chi Aurigse in January 1892. It was discovered with a hand telescope by Dr. T. A. Anderson of Edinburgh. When at its brightest it was visible to the naked eye. ASTRONOMY WITHOUT A TELESCOPE. 247 the naked eye than with any instrument, although an opera-glass brings out well, in some places, its more delicate details. A mere passing glance might lead a casual observer to suppose that the Galaxy stretched as a band of nearly uniform brightness across the heavens. But good eyesight, careful attention, and a clear sky will soon disclose numerous details pre- viously unsuspected ; streams and rays of different brightness, intersected by rifts of darkness, and inter- spersed with spots and channels of comparatively starless spaces. An excellent drawing of the Milky Way the result of five years' observations with the naked eye alone has recently been completed by Dr. Otto Boeddicker, at Lord Rosse's observatory in Ireland. This beautiful picture is exquisitely drawn, and shows a wonderful amount of detail. A writer in the Saturday Review of November 30, 1889, says, " His maps are in many respects a completely new disclosure. Features barely suspected before, come out in them as evident and persistent; every previous representation appears, by comparison, structureless" This shows what can be done with the naked eye in the study of this wonderful zone. Among the Nebulae and clusters there are not many objects visible to the naked eye. A hazy appearance about the middle star in Orion's " sword " indicates the presence of the " great Nebula," one of the finest objects in the heavens. The "great Nebula in Andromeda," aptly termed "the Queen of the 248 THE WORLDS OF SPACE. Nebulae," is distinctly visible to the naked eye on a very clear night. It lies near the four and a half magnitude star, Nu Andromedae (a few degrees north of Beta Andromedae), and may be well seen in the early evening hours in the month of January, when it is high in the sky. It somewhat resembles a small comet. This Nebula was known long before the invention of the telescope, and it was described by one of the earlier astronomers as resembling " a candle shining through horn," a not inapt description. Of star clusters visible without optical aid may be mentioned the double cluster Chi Persei, which appears to the eye as a luminous spot in the Milky Way; the cluster known as 35 Messeir, a little north of Eta Geminorum, just visible to the naked eye on a very clear night ; and there are others in the Southern Hemisphere, notably the globular cluster known as Omega in the Centaur, which shines as a hazy star of the fourth magnitude. Among the clusters may perhaps be included the Praesepe, or the " Bee-hive/' in Cancer, which has a nebulous appear- ance to the naked eye. Coming now to the Solar system, the Sun and Moon, of course, first attract attention. Cases of sun-spots visible to the naked eye are recorded, but of course spots of such enormous size are of rare occurrence. 1 Of lunar detail, but little can be seen 1 A remarkable spot of this kind was visible in February 1892, and another in February 1894. ASTRONOMY WITHOUT A TELESCOPE. 249 without a telescope of some sort, but the larger markings are sufficiently distinct to good eyesight to convince the observer that they do not alter percep- tibly, thus showing clearly that the Moon always turns the same side to the Earth. Of the planets, nothing of their appearance in the telescope can, of course, be seen with the naked eye, but it is easy to identify the brighter planets. Mercury, owing to its proximity to the Sun, is rarely visible in this country, but, when favourably situated, it may sometimes be detected near the Sun shortly after sunset, or a little before sunrise. Notwithstand- ing the difficulty of seeing it, it was well known to the ancients, an observation of the planet dating back to 264 B.C. It is easier, however, to see in more southern latitudes, and I have frequently observed it as bright as a star of the first magnitude in the clear air of the Punjab sky. I have also seen it on several occasions in Ireland, and the Rev. S. J. Johnson, F.R.A.S., tells me he has seen it with the naked eye no less than 100 times in the south of England. The brilliant planet Venus can hardly be mistaken when seen in the morning or evening sky. When at its brightest it considerably exceeds Jupiter and Mars, and far surpasses Sirius, the brightest star in the heavens. If a very bright planet is seen rising at sunset, it cannot be Venus, which is never seen beyond a limited distance from the sun. The observer may, therefore, 250 THE WORLDS OF SPACE. conclude with certainty that the planet is either Jupiter or Mars. The latter, which occasionally rivals Jupiter in brilliancy, may be easily distin- guished from the "giant planet" by its distinctly reddish colour. Saturn shines with a yellowish light, and is never so bright as Mars or Jupiter when at their brightest. The planet Uranus is just visible to the naked eye, and may be found without optical assistance, when its position is accurately known. At present (1894) it lies near the star Alpha Librae. Some observers think that they can see the crescent of Venus with the naked eye when the planet is in that phase, but this seems very doubtful. Cases have been recorded of one or two of the satellites of Jupiter having been seen with the unaided eyesight, but few are gifted with such keen vision. Occultations of bright stars may be well seen with the naked eye, especially when they pass behind the moon's dark limb, and as the disappearance of a star is practically instantaneous, really valuable observa- tions may be made without a telescope, by merely noting the exact time at which the star vanishes. Most of the comets discovered by astronomers are small and faint, and only visible in good telescopes. At intervals, however, a brilliant visitor appears on the scene, and its path among the stars may be watched from night to night with the naked eye. Before the invention of the telescope, bright comets were watched in this way, and their course recorded ASTRONOMY WITHOUT A TELESCOPE. 251 so carefully, that it has been found possible to calcu- late their orbits with some approach to accuracy. In these days of large telescopes and instruments of almost mathematical precision, such a method of observation is of course superseded ; but we may still watch the movements of a bright comet with interest, and note its apparent path across the sky with pleasure and profit. Shooting-stars and fireballs may be best observed with the naked eye, and the excellent work done in this way by Mr. W. F. Denning, F.R.A.S., should encourage others to take up this interesting branch of astronomy. Another object which may be well seen with the naked eye indeed it may be best observed in this way is the Zodiacal Light. This is a lenticular or cone-shaped beam of light, which makes its appear- ance at certain times of the year, above the Eastern horizon before the dawn, and above the Western horizon after sunset, when the sky is clear and the moon absent. In the tropics it is much more easily seen, the twilight being shorter, and I have often observed it in India shining with great brilliancy. But even in this country useful observations may be made of its position among the stars, its brilliancy relative to the Milky Way, and other details. I have often observed it in the west of Ireland, and have sometimes seen it exceeding in brightness the Galaxy between Cygnus and Cassiopeia. 252 THE WORLDS OF SPACE. From the above sketch my readers will see how much may be learned of astronomy without optical assistance of any kind, and I hope that those who do not possess a telescope will use their eyes instead, and thus gain some knowledge of the wonders and beauties of the starry heavens. The knowledge thus acquired will stimulate their curiosity, and will give them a keener interest in reading books which describe the still greater wonders revealed by the telescope. XXV. RECENT ADVANCES IN ASTRONOMY. THE study of astronomy is now making rapid strides. Greater progress is made in these days in a month than was achieved in former times in a year. Discovery now follows discovery in such quick suc- cession, that books on astronomy very soon become obsolete and unreliable in their data, and it is only by the frequent issue of new editions that even popular works on the subject can keep up with the ever-growing mass of new knowledge which is being almost daily added to our store. Mere suspicions of yesterday become certainties to-morrow, and discoveries almost undreamt of occasionally startle even those whose knowledge of astronomy is some- thing more than elementary. This rapid advance in astronomical knowledge is chiefly due to two causes. First, the greatly increased interest now taken by the general public in the noble science, which has justly been called "sublime." Amateur astronomers have largely multiplied in number in recent years, and many of these are doing THE WORLDS OF SPACE. excellent work, even with small instrumental means. A remarkable instance of the interest taken in astronomy is afforded by the success attained by the British Astronomical Association. Started in the autumn of 1890, the number of its members had reached 800 before the close of 1893, a fact probably unique in the history of scientific societies. Other astronomical societies have also been recently founded, and the interest in the science seems to be rapidly spreading. This increased interest and desire to know more of the wonders of the heavens has naturally led to the second cause of the rapid increase of astronomical discoveries, namely, the construction of larger telescopes, and the invention of new methods of research. Some 50 years ago a refractor of 10 inches aperture was considered a large telescope, but now we have instruments with object glasses of 2 5> 3O, and 36 inches aperture, and the construction of even larger telescopes is in contemplation. The application of photography to the heavenly bodies has caused quite a revolution in the study of astronomy. This is due to the introduction of the sensitive dry plates now generally used by photo- graphers. Most successful results have been obtained in celestial photography by the Brothers Henry, at the Paris Observatory, by Dr. Common and Dr. Roberts in England, and others. An ordinary portrait lens strapped to a telescope driven by clockwork has also been successfully employed in RECENT ADVANCES IN ASTRONOMY. 255 photographing portions of the Milky Way, by Mr. Barnard, at the Lick Observatory, California ; Mr. Russell, at the Sydney Observatory ; and by Dr. Max Wolf, at Heidelberg. These are of great beauty, and give promise of still greater results in the near future. The stellar photographs taken at the Paris Observatory are especially beautiful, the stars appearing as small circular discs on a black background. Star clusters, nebulse, and the great globular cluster in Hercules (13 Messier) have also been successfully photographed. Fine photographs of the great nebulae in Orion have also been obtained by Dr. Common, Dr. Roberts, and Mr. Russell. These show a great extension of nebulous light. A wonderful photograph of the great nebula in Andromeda (31 Messier), by Dr. Roberts, shows that the " dark lanes " originally observed by Bond represent spaces between rings, into which the nebula is divided. These rings suggest a stage in the evolution of a stellar system on the lines of Laplace's Nebular Hypothesis. . Photographs of the Pleiades taken at the Paris Observatory, and by Mr. Roberts in England, show that all the brighter stars of the groups are involved in nebulous light. No less than 2326 stars can be counted on the Paris photograph. The great nebula surrounding the curious variable star, Eta Argus, in the Southern Hemisphere, was carefully observed by Sir John Herschel, during his 256 THE WORLDS OF SPACE. residence at the Cape of Good Hope in the years 1834-38. He gives an elaborate and beautiful draw- ing of the nebula in his Cape Observations. The nebula has since been frequently observed by other astronomers, and suspicions have arisen that great changes have taken place in its appearance since HerschePs time. A photograph of the nebula has recently been obtained by Mr. Russell at the Sydney Observatory, which shows that " one of the brightest and most conspicuous parts of the nebula/' near the centre of Herschel's drawing, has "wholly dis- appeared ! " and its place is now occupied by " a great dark oval." Mr. Russell also finds that the portion referred to is now invisible in a telescope of II J inches aperture. That the photograph shows more than Herschel's telescope could possibly have done is proved by the fact that in the " lemniscate " or "key-hole" opening in the nebula, where Herschel saw one star, the photograph shows ten ! The vanished portion of nebulous light was first missed by Mr. Russell in the year 1871, and the photographs now prove the correctness of his earlier observation. The accuracy of Sir John Herschel's drawing is shown by the fact that his descriptive paragraphs of other parts of the nebula "accurately describe the features as they are to-day," and leave no room for doubt that a remarkable portion of the nebula has actually disappeared since the date of his drawing. Mr. Russell has also photographed the " Magellanic THE NEBECULA MINOR. Taken by H. C. RUSSELL. October 14^15^, 1890. THE NEBECULA MAJOR. Taken by H. C. RUSSELL, October ijt/t, 1890. 4 \ RECENT ADVANCES IN ASTRONOMY. 257 clouds." The larger cloud, the Nubecula Major, as it is called, he finds 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. In these wonderful objects numerous small stars are seen to be mixed up with nebulous light. A photograph of the well-known " coal-sack," near the Southern Cross, shows that it is not completely devoid of stars, as its appearance to the naked eye might lead us to suppose. Numerous faint stars are visible in the photograph, and only near its northern border is there a spot completely free from stars. Photographs of the well-known Milky Way feature, "the gap in Argo," show that this apparent break in the course of the Galactic stream is "as thickly studded with stars and star dust as the neighbouring parts figured on the best star maps." It would therefore seem that the breaks and vacuities visible to the naked eye in the Milky Way are more apparent than real, and these photo- graphs afford further evidence against the "cloven disc " theory of the Milky Way. At the International Congress of Astronomers, re- ferred to in a previous chapter, it was decided that a photographic chart of the whole heavens should be undertaken at several observatories, a portion of the sky being allotted to each. These charts will include all stars to the fourteenth magnitude. The prelimin- ary arrangements for carrying out this plan are now 258 THE WORLDS OF SPACE. nearly completed, and it is hoped that in a few years a complete chart of the heavens as they existed at the close of the nineteenth century will be avail- able for the study of future generations. The invention of the spectroscope, and its applica- tion to the study of celestial bodies by Dr. Huggins, has added greatly to our knowledge of the chemical constitution of the stars. By means of this wonderful instrument of research we can now analyze the light emitted by the Sun and stars, and by a comparison with the spectra of terrestrial substances we can ascertain the chemical elements which exist in the glowing photospheres of these distant orbs. A careful examination of stellar spectra has enabled astronomers to divide them into five types. The first of these shows the hydrogen lines abnormally dark. This type of spectrum is shown by Sirius and numerous other stars. The second type corresponds closely with the solar spectrum, in which many fine lines are visible, denoting the presence of metallic elements in the gaseous state. The third and fourth types include those in which dark bands are visible instead of fine lines. The number of stars with spectra of this type is comparatively small. They are mostly of an orange or red colour, and their chemical constitution is still somewhat uncertain. The fifth type includes the planetary nebulae and some rare stars showing bright lines in their spectra. It was discovered by Dr. Huggins that the motion of a star in the line of RECENT ADVANCES IN ASTRONOMY. 259 sight, that is towards, or away from, the Earth, could be ascertained by observations of the displacement of the spectral dark lines when compared with the spectrum of a terrestrial substance such as hydrogen. Observations of this kind made at Greenwich Observ- atory and elsewhere gave very discordant results, owing to the extreme difficulty of accurately making such delicate measurements. This difficulty has now been overcome by measurements of the photographed spectra. Remarkable results have been obtained in this way by Professor Vogel, at Potsdam, the velocities of approach and recession being ascertained to a quite unexpected degree of accuracy. The measure must, of course, be corrected for the effect due to the Earth's orbital motion round the Sun, and so reliable are the results obtained from the photographs, that Mr. Keeler, the well-known American astronomer, has stated that he would undertake to determine the month of the year if he did not know it by spectro- scopic observations of the great nebula in Orion ! Indeed, it has been suggested that when this method of research has been further developed, it may be possible to reverse the process, and to ascertain from stellar spectra the earth's velocity in its orbit at any time, and hence the Sun's distance from the Earth, by a novel method. A catalogue of the spectra of over 10,000 stars has recently been formed at the Harvard Observatory (U.S.A.). It is known as the Draper Catalogue, 260 THE WORLDS OF SPACE. and was undertaken as a memorial to the late Dr. Henry Draper, who did much good work in this branch of astronomical research. The Catalogue was constructed from an examination of photographs of the stellar spectra, taken with a telescope of 8 inches aperture, having a focal length of 44 inches. For the brighter stars an exposure of at least five minutes was given to each plate, which con- tained two to four regions, 10 square. For the fainter stars an exposure of one hour was allowed to each plate of 10 square. The Catalogue includes stars down to 25 south declination. A more de- tailed examination of the spectra of the bright stars has also been made with the n-inch telescope and 15-inch equatorial of the same Observatory. This latter research has led to some remark- able results. In the case of the bright stars Zeta Ursae Majoris (Mizar) and Beta Aurigae, the more conspicuous of the spectral lines are seen to be doubled periodically, indicating that these stars con- sist of two components revolving round each other, and so close that the highest powers of the largest telescopes are unable to separate them. Indeed, Professor Pickering has shown that " the effect of the prism may be regarded as multiplying the magni- fying power of the telescope about 5000 times." Knowing the orbital velocity, we can easily compute the absolute dimensions of the stellar system in miles, and hence its mass in terms of the Sun's mass becomes RECENT ADVANCES IN ASTRONOMY, 26 1 known. By this new method we can, therefore, calculate the mass of a star the distance of which from the Earth is unknown ! x Professor Pickering has recently discovered that the majority of the brighter stars in the Milky Way show spectra of the first or Sirian type, and, judging from the facility with which the smaller stars of the Galaxy impress their image on the photographic plate indicating bluish light he concludes that the fainter stars, which by their clustering produce the nebulous light of the Milky Way, have also, most probably, spectra of the Sirian stamp. From a careful enumeration of the stars in the Draper Catalogue, which lie on the Milky Way and its branches, as drawn by Heis, I find that out of 3061 stars, 1940 are of the first type, 1 100 of the second or solar type, and the remainder of the third, etc. This gives a per- centage of over 63 for stars of the Sirian type a remarkable result. Possibly it may some day be shown that the rule is a general one, and that all the stars of the Milky Way have spectra of the first type, those showing spectra of other types being merely projected visually on the Galactic zone. If this be so, we may conclude that the Milky Way is not con- stituted in the same way as the rest of the sidereal universe, and has probably had a different origin another argument against the "disc" theory. It was long since suggested that the periodical 1 See chapter on ' Weighing the Stars.' 262 THE WORLDS OF SPACE. diminution of light in the famous variable star Algol was possibly due to the interposition of a dark eclipsing satellite. The correctness of this hypothesis has now been proved by a photographic examination of its spectrum by Professor Vogel, at Potsdam. He finds that before the minimum of light takes place, the brighter component is receding from the Earth, and after the minimum is past, it is approaching. In this case, as one of the components is a dark (or nearly dark) body, the spectral lines are merely shifted slightly in position, not doubled as in the case of Beta Aurigae and Mizar. In these latter stars it appears that the plane of motion does not pass through the Earth, for, otherwise, there would be some decrease in the star's light when one component passed in front of the other. No such variation of light has been detected in either Mizar or Beta Aurigae. Discoveries in other branches of astronomy have also made rapid progress in recent years. The num- ber of double stars has been largely increased, chiefly by the labours of that excellent observer, Mr. S. W. Burnham. Many of the interesting objects discovered by this accomplished astronomer already show signs of orbital motion, and the calcu- lation of their orbits will, in the course of time, afford ample work for future computers. A number of orbits of binary stars have been recently calculated, and the number of these remarkable systems, for which more or less accurate orbits have been com- RECENT ADVANCES IN ASTRONOMY. 263 puted, now amounts to about 80. The faint com- panion of Sirius, which revolves round this brilliant star in a period of 50 to 60 years, has now approached so close to its primary, that it has passed beyond the reach of even the giant telescope of the Lick Observatory. It will, however, in a few years more, emerge from the rays of the bright star, and it will then be possible to compute its orbit with greater accuracy. The number of known variable stars is also steadily increasing, and the fact that some of the long period variables show bright lines in their spectra, has been taken advantage of to detect some new objects of this class with the spectroscope. It should be noted, however, that large telescopes and photographic apparatus are not necessary for the discovery of variable stars, many of them having been found with an opera-glass or binocular. Four new variables have been added to the list by the present writer, by means of an ordinary binocular field-glass, similar to those used on race-courses. Variables of the Algol type are very rare, only ten altogether having been hitherto detected. As, however, the light of these variables only fluctuates for a short time near the minimum, the difficulty of detection may very possibly account for the small number known. Professor Schiaparelli's observations of curious markings on the surface of the planet Mars, which he terms " canals," have been partially verified by 264 THE WORLDS OF SPACE. other observers. Some, however, doubt their objec- tive existence. The observations made at the opposition of 1892 are rather conflicting, and do not add very much to our knowledge of the planet. Schiaparelli's supposed discovery that the rotation periods of Mercury and Venus are the same as their periods of revolution round the Sun has not yet been fully confirmed by other observers. Indeed, in the case of Venus, M.M. Niesten and Stuyvaert, of the Brussels Observatory, find that a period of about 23^- hours, originally found by de Vico, will suit their observations very well. The observations are, how- ever, of great difficulty, owing to the faintness of the markings visible on these planets, and the peculiar position of the planets themselves with reference to terrestrial observers. The discovery of minor planets between Mars and Jupiter still continues. Only 13 of these small planets were known in the year 1850. The number reached 100 in 1868, 200 in 1879, 300 in 1890, and now (Feb. 1894) over 380 have been discovered. It would seem that the number of these small planets is probably very large, and it will soon become a difficult matter to keep an accurate account of them. For instance, a supposed new planet found by Palisa on August I4th, 1891, afterwards proved to be iden- tical with Medusa (No. 149), which was discovered by Perrotin in September 1875, and several other similar cases have recently occurred. RECENT ADVANCES IN ASTRONOMY. 265 On September 9, 1892, Prof. E. E. Barnard, ob- serving with the 36-inch refractor of the Lick Observ- atory, made the remarkable and unexpected dis- covery of a fifth satellite to the planet. It is nearer to Jupiter than the other known satellites, and revolves round the planet in a period of II hours, 57 minutes, 22\ seconds, at a mean distance of 112,500 miles from the centre of its primary ; or about 67,000 miles from the surface of Jupiter. The new satellite is a faint object even in large telescopes, and its diameter does not probably exceed 100 miles. The supposed duplicity of Jupiter's first satellite, suspected by Messrs. Barnard 'and Burnham at the Lick Observatory, in September 1890, has not been confirmed. Mr. Barnard now considers the appear- ance as probably due to a white belt on the surface of the satellite, nearly parallel to the belts of Jupiter. When the satellite transits a bright portion of Jupiter's surface, this bright belt becomes invisible, and the other portions of the satellite appear like two dusky spots, thus giving the satellite the appear- ance of being double. Further observations will be necessary to prove the accuracy of this hypothesis, but that the satellite is really double would seem very improbable. Later observations of this satellite by Professors Campbell and Schaeberle of the same Observatory, show that it is " ellipsoidal, with one of its longer axes directed towards the centre of Jupiter/' An interesting observation of an eclipse of Saturn's 266 THE WORLDS OF SPACE. satellite, Japetus, in the shadow of the globe and rings of the planets, was made by Mr. Barnard at the Lick Observatory on November 1st, 1889. He says " The observations show that after passing through the sunlight shining between the ball and rings, Japetus entered the shadow of the crape ring. As it passed deeper into this, the absorption of sunlight became more and more pronounced, until finally the satellite entered the shadow of the bright rings. In a word, then, the crape ring is truly transparent, the sunlight sifting through it. The particles composing it cut off an appreciable quantity of sunlight. They cluster more thickly, or the crape ring is denser, as it approaches the bright rings." From the total dis- appearance of the satellite when it passed behind the bright ring, he concludes that " the bright ring is fully as opaque as the globe of Saturn itself." 1 1 Monthly Notices, R.A.S., January 1890. XXVI. SOME ASTRONOMICAL ERRORS AND ILLUSIONS. ALL books contain some errors either misprints, or mistakes due to oversights, or imperfect knowledge of a special subject. No man's work is perfect. Misprints seem sometimes unavoidable, even when the greatest care is taken in the correction of proof sheets ; but these are easily detected, and are not of so much importance as errors in the statement of facts and principles. These latter are of course very misleading to the student of astronomy, and in the interests of truth should not, I think, be allowed to pass uncorrected. It may, perhaps, seem rather in- sidious to call attention to mistakes in other men's work ; but as some of the errors referred to, unfortu- nately, occur in my own writings, I have less hesita- tion in making the necessary corrections, the accuracy of which may be tested by reference to the authorities quoted. Proctor's works are generally very accurate, and free from error. I have, however, found a few trivial mistakes, which, although they do not detract from 268 THE WORLDS OF SPACE. the value of his excellent writings, should, I think, be corrected. In the article on ( Coloured Suns/ in his valuable Essays on Astronomy, he speaks of the components of the beautiful double star c Bootes as " nearly equal" in brightness. 1 In reality, however, they are rather unequal, the magnitudes being about 3 and 7, 2 a difference which implies that the primary star is nearly forty times brighter than its companion. In his article on * A Novel Way of Studying the Stars,' in the same volume, Proctor estimates the area of the heavens covered by the Milky Way as one-tenth of the Northern Hemisphere, and one-eleventh of the Southern. This may be correct enough for the very diagrammatic representation of the Galaxy given in most popular star atlases ; but in Heis's more care- fully drawn delineation, I find that the Milky Way covers about one-fourth of the Northern Hemisphere. The southern half of the Galaxy, as drawn by Dr. Gould in the Uranometria Argentina, is of still greater extent, covering, according to Markwick, about one- third of the Southern Hemisphere. In his interesting little work Half-Hours with the Telescope, Proctor says that all four satellites " are visible in a good opera-glass " ; 3 but probably they could only be seen with such an instrument by those gifted with exceptionally keen eyesight. In 1 The star is, however, correctly described in his Half-Hours ivith the Telescope. 2 Webb's Celestial Objects, 4th edition, p. 243. 3 p. 85. SOME ASTRONOMICAL ERRORS AND ILLUSIONS. 269 any case, atmospheric conditions would have much influence on their visibility, even in a powerful binocular. In describing Orion in the same volume, he speaks of Betelgeuse (a Orionis) " as one of the most remarkable variables in the heavens." This is, however, incorrect, as the star remains for long periods constant in light, or nearly so ; and even to a careful observer of variable stars, it often continues for months together without perceptible change. Its variation, even when most conspicuous, is smallj and might easily be overlooked by an ordinary observer. It was recorded by Hevelius that the famous vari- able star Mira (o) Ceti was invisible in the years 1672 to 1676. From this it was inferred that the star did not rise to a maximum in those years, and this statement has been repeated in several text-books of astronomy. The idea is, however, quite incorrect ; for it was long since (1837) pointed out by Bianchi 1 that the supposed invisibility of Mira in the years referred to by Hevelius was simply due to the fact that the maxima in those years occurred at a time when the star was near the Sun, and could not be observed. In fact, as the period from maxi- mum to maximum is about 331 days, eleven periods will be nearly equal to ten years. Hence if the star, in any particular year, happens to be at a maximum when near the Sun (in April), it will be in nearly the same position when at maximum 1 Astronomische Nachrichten. 270 THE WORLDS OF SPACE. after an interval of ten years. Hence, roughly speak- ing, the star's maximum will be invisible every ten years. Of course for a year or so before and after this time it will be too close to the Sun for observa- tion ; and hence the maxima will pass unnoticed for several years at least to ordinary observers. The non-appearance of Mira in the years 1672 1676 is therefore satisfactorily explained. In several books on astronomy it is stated that Mira wholly disappears at minimum. This, however, is quite incorrect. The star never descends below magnitude 9^, and when favourably placed for observation, may always be seen in a telescope of three inches aperture. 1 With reference to the variable star Algol, it is sometimes stated that the period is increasing ; but, on the contrary, it is still diminishing, as Chandler's investigations have clearly shown. It is also fre- quently stated that the variation of light is from mag- nitude two to magnitude four ; but at its brightest, the star is always a little fainter than an average star of the second magnitude, and at minimum it never descends so low as the fourth magnitude. The total variation certainly does not exceed one and a half magnitude, and Professor Pickering's photometric measures at Harvard College show that it does not much exceed one magnitude. My own observations agree with Professor Pickering's result. The remarkable variable x Cygni has been often 1 Chandler's Catalogue of Variable Stars. SOME ASTRONOMICAL ERRORS AND ILLUSIONS. 2/1 confused with FlamsteecTs x r 1 7 Cygni ; but the latter is quite a distinct star. The variable is the true x Cygni of Bayer, and Flamsteed affixed the letter x to his number 17 of Cygnus by mistake, the variable having been faint at the time of his observa- tion. Stone proposed to call 17 x 1 ) an d the variable X ' 2 ', but there seems to be no necessity to perpetuate an error which has long since been pointed out. All authorities on variable stars now give this variable its correct designation x Cygni. With reference to the apparent motions of the planets in the sky, it is stated in some works on astronomy that a planet is stationary when it is moving in a line directed towards an observer on the Earth. Professor Lockyer says, " the planet, as seen from the Earth, will appear at rest as we are advancing for a short time straight to it " ; but this is quite incorrect ; on the contrary, when in this position it has a rapid direct motion. 1 Herschel and other astronomers give the correct explanation of the stationary points. With reference to the apparent motion of the binary or revolving double stars, as seen from the Earth, the idea seems a not uncommon one that the minimum distance between the components always occurs when the companion star is at the periastron of the real orbit. Some years since a well-known astronomer expressed his opinion that the famous southern binary star a Centauri had not then reached the 1 Sidereal Messenger ', May 1889, p. 233. 2/2 THE WORLDS OF SPACE. periastron, because the components had not yet arrived at their point of nearest approach. But this is quite a mistake. As a general rule, the minimum distance does not occur at the periastron point. Under certain conditions, of course, the minimum apparent distance may coincide with the periastron, but usually this is not the case. From my experience in the calculation of binary star orbits, it will, I think, be admitted that I am quali- fied to speak with some authority on this point ; and I feel confident that my contention will not be disputed by any one who really understands the subject. In some books on astronomy it is stated that the temporary star of 1866 in Corona Borealis the " Blaze Star," as it is sometimes called was discovered simultaneously by several observers ; but as is well known, the star was really discovered by the late Mr. Birmingham, at Tuam, Ireland, at midnight on the night of May 12 in that year. In his work, LOrigine du Monde, M. Faye, the French astronomer, has the following misleading statement "Le cas le plus frappant et le mieux observe est celui de 1866. M. Courbebaisse, ingenieur des pont est chaussees, vit un soir (13 mai) briller dans la Couronne boreale une belle etoile qu'il n'avait pas remarquee les jours precedents," and he does not mention Mr. Birming- ham ! That the star was seen on the night of May 13 may perhaps be true; but it was certainly first SOME ASTRONOMICAL ERRORS AND ILLUSIONS. 2/3 seen by Mr. Birmingham on the preceding night, and to him all authorities on variable stars unanimously award the discovery. This is not a question of Adams or Leverrier, and Neptune ; the credit of the discovery is undoubtedly due to the Irish astronomer, and to him alone. In the fourth edition of The Heavens, by Guillemin, it is stated with reference to Betelgeuse, that this bright star " has recently descended to the sixth mag- nitude." 1 This is, however, quite erroneous. The star although certainly variable to a small extent has never, so far as I know, been observed as faint as even the second magnitude. The extreme range of variation does not probably exceed half a mag- nitude. I have never seen it myself fainter than Alde- baran, and it usually exceeds that star in brightness. I will now notice some errors in my own writings. In my Planetary and Stellar Studies it is stated that Tycho Brahe's star of 1572 "increased rapidly in brilliancy " 2 after its discovery ; but this would seem to be a mistake. The star more probably appeared suddenly at \\.sfull brightness. In p. 148, the discovery of the temporary star of 1604 is attributed to Mostlin ; but it was really discovered by Brunowski, a pupil of Kepler's, on October 10 of that year. Mostlin was Kepler's old master, and also observed the star, but did not see it well till October 16. In the chapter on ' Stellar Photography ' it should have been stated l p. 271. 2 p. 147- 2/4 THE WORLDS OF SPACE. that the "three separate stellar photographs" 1 are taken on the same plate, the plate being slightly shifted at the second and third exposures, so as to make each star finally appear in three small discs, forming an equilateral triangle. In the note at p. 244 of the same volume, in the passage, " the great inequality in the motion of Jupiter and Saturn depends on the exact equality, etc.," the word should be approximate, not " exact." In The Scenery of the Heavens* it should have been stated that Secchi was probably mistaken when he thought he could see bright lines in the spectrum of the variable R Geminorum. A subsequent careful ex- amination of the star's light by Vogel showed the spectrum to be really due to carbon, and not to hydro- gen. At p. 117, the discovery of the great comet of 1680 is ascribed to Halley, whereas G. Kirch was the real discoverer. On p. 44 the length of the " Hyginus cleft" on the Moon is given as nine miles instead of ninety ; and there are a few other misprints. Many telescopic observations have been made from time to time even by skilled observers which other observers have been unable to verify. The weight of evidence against the reality of these observations is so strong that we are compelled to consider them as optical illusions, due either to some peculiarity in the observer's eye, or to some defect in J p. 213. 2 p. 258. SOME ASTRONOMICAL ERRORS AND ILLUSIONS. 275 the telescope employed. A short account of some of the most remarkable of these cases may prove of interest to the general reader. From irregularities in the observed " proper motion " of the bright star Procyon, astronomers came to the conclusion that probably its motion was disturbed by the attraction of a close companion as in the well- known case of Sirius. In 1873 tne great Russian astronomer, Otto Struve, believed that he could see a faint companion to this brilliant star. Every pre- caution was taken to prevent illusion, and to test the accuracy of the discovery. The position of the faint star was even measured by Struve and his assistants ! In 1874 the companion was again observed by Struve, and its position was found to agree closely with that deduced by Dr. Auwers from theoretical consider- ations. Other Russian astronomers thought they could see it also, and it was even measured by one observer in England. The Washington observers, however, failed to see it with their great 26-inch refractor. At last Struve himself announced that his supposed companion was merely an optical illusion, or "ghost," as these imaginary telescopic stars are called by astronomers. Mr. Burnham examined Procyon in October 1888, with the great Lick telescope, and says, " Carefully examined with all powers up to 3300 on the 36-inch under favour- able conditions. Large star single, and no near companion." 276 THE WORLDS OF SPACE. In 1863, Gold schmidt, observing with a telescope of only 4 inches aperture, announced the discovery of five faint stars within i' of arc of Sirius. Dawes saw one of these, but failed to see any of the others. Secchi, however, in 1865, found a companion at 44" distant, and Marth, at Malta, another at about 70". Burnham with the Lick telescope found the well- known close companion a " very easy object " in 1888, and adds, " I have carefully looked for other stars near Sirius, but without finding anything worth noting." Some of Goldschmidt's companions must therefore have been "ghosts." The Pole star has a small companion of the ninth magnitude, which forms a well-known test for small telescopes. In the year 1869, M. A. de Boe, a well- known Belgian astronomer, announced his discovery of two faint stars nearer to the bright star than the companion. Other observers thought they could also see these faint attendants, but the great telescope at Chicago failed to reveal their existence to Burnham, who said " I have no hesitation in saying these supposed stars do not exist." Observing with the giant telescope of the Lick Observatory in April 1889, Burnham says " Carefully examined with the 36-inch with all powers. Both stars single, and no companion nearer than the Struve star." We may therefore safely conclude that M. de Boe's supposed stars were mere optical illusions. A supposed companion to the fourth magnitude SOME ASTRONOMICAL ERRORS AND ILLUSIONS. 277 star, 72 Ophiuchi, probably forms another example of a telescopic " ghost." It was discovered as a double star, or supposed double star, by Otto Struve, on November i, 1841, the companion being rated of the seventh magnitude, but it was seen single on May 14, 1842. It was again seen double in September 1842, but was observed to be single in the years 1844, 1848, 1850, 1851, 1852, and 1859. In 1876, however, Struve again saw the companion very distinctly, and mea- sured its position with reference to the brighter star. He concluded that the star was really double, but that the companion was subject to great fluctu- ations of light. It was measured by Madler in 1845 and 1847, and in 1859 Secchi found the components well separated only three weeks before it was found single by Struve at Poulkova ! On two very fine nights in 1874, Professor Newcomb could not see any trace of the companion with the 26-inch refractor of the Washington Observatory, and Professor Hall was equally unsuccessful in 1876. Burnham found it "certainly single" on "a first-class night" in 1880, with an i8J-inch refractor, and observing the star with the great Lick telescope in April 1889, he says, " Large star, single, with 36-inch, and no near companion." Sir William Herschel, observing Saturn in the year 1 805, thought that the ball of the planet was not only flattened at the poles as it is known to be but also occasionally at the equator, giving the planet "a 278 THE WORLDS OF SPACE. "square-shouldered" aspect. Proctor attempts to explain this appearance by an optical illusion, pro- duced by the different curvature of the ball and rings, but admits that the phenomenon may possibly have an objective reality due, perhaps, to physical changes in the planet's atmosphere, which is probably of con- siderable depth. As Herschel verified his observations with different instruments and by actual measurement, and as the same phenomenon has been noticed by other observers, it does not seem quite certain that it was merely due to an optical effect. Sir W. Herschel thought he could see six satellites to the planet Uranus. These, with the two inner ones, Ariel and Umbriel, discovered in 1847, would make eight in all, but the large telescopes of modern times fail to show more than four. Some of those seen by Herschel must, therefore, have been either optical "ghosts," or else small fixed stars which happened to lie near the planet's path at the time of observation. Herschel also suspected, at one time, that he could see traces of rings round Uranus like like those round Saturn, but his observation was never confirmed, either by himself or other observers, and we now know that this planet has no such appendage. Lassell and Bond thought they could see traces of a second satellite to Neptune, but this suspicion has never been confirmed. When Venus is in the crescent form, or near inferior SOME ASTRONOMICAL ERRORS AND ILLUSIONS. 279 conjunction, many observers have thought that they could see the dark portion of the disc in the same way that the dark part of the Moon is visible when only a few days old, " the old Moon in the young Moon's arms," as it is popularly called. This phenomenon is easily and satisfactorily explained in the case of the Moon by sunlight reflected from the Earth, but is so unaccountable in the case of Venus, as a moments consideration will show, that some astronomers have attributed the appearance to an optical illusion. The phenomenon was, however, observed so far back as 1715, and it has often been seen in recent years by such excellent observers as Browning, Elger, Erck, Franks, Grover, Webb, and Winnecke. It was seen by Franks and Perkins in 1884; darker than the sky by the former observer, and brighter than the sky by the latter. Another observer, Mr. J. M. Offord, described it as of" a prussian-blue colour.' 1 It is even said to have been seen in the daytime by Andreas Mayer in October 1759, and by Winnecke in 1871. As an attempt at a solution of the mystery, a com- parison has been suggested with some of our terrestrial nights, which are much brighter than others, due to a phosphorescent glow over the whole sky, which has been noticed by Arago, Schroter, Webb, and by the present writer both in this country and in India. It must be admitted, however, that the hypothesis is not a plausible one, and does not give a very satisfactory explanation of this puzzling phenomenon. 280 THE WORLDS OF SPACE. Several observations have been recorded of a supposed satellite of Venus, but it has now been clearly proved that these were due either to imperfec- tions in the telescope used, or to small stars near which the planet passed in its course. XXVJI. THE ARITHMETIC OF ASTRONOMY. IN works on astronomy, numerical results are usually given with reference to the distances of the Sun, planets, and fixed stars, their diameters, densities, etc., but no attempt is made at least in popular works to explain to the general reader the method of calculation by which these results have been arrived at. On making inquiry from those qualified to give information on the subject, the reader will probably be told that the calculations are too complicated to admit of popular explanation, but that the figures may be accepted as undoubtedly correct. This may possibly satisfy the reader who is not versed in figures, but to those who have received at least the elements of a mathematical education, the reply will not be a satisfactory one. They will probably desire to know something more of the methods by which the results have been obtained, or at least how the figures given in the text-books have been found from the measurements derived from the observations. It has, therefore, occurred to me that an elementary 282 THE WORLDS OF SPACE. account of the methods by which these figures are found may prove of interest to those who are familiar with the elements of mathematics. I do not propose to describe here the methods of observation employed to determine the numerical results, but merely to show how these figures are found from the observ- ational data. To begin with the most important of these numbers, the Sun's distance from the Earth, we often hear of the Sun's "parallax," that its probable value is 8*8" of arc, and that the resulting mean distance of the Sun is nearly 93,000,000 miles. What is the meaning of these statements ? I will not here enter into the methods of observation by which the value of the " solar parallax " is measured. But, suppos- ing this parallax to be 8'8", what is the meaning of the statement, and how are we to obtain from it the Sun's distance from the Earth ? Those who are familiar with the first principles of Trigonometry will know that the angle subtended by an arc equal in length to the radius of the circle is 206,265". One foot, therefore, on the circumference of a circle of 206,265 feet radius, will subtend an angle of i" at the centre of the circle, and 8'8 feet an angle of 8'8". Now the Sun's "parallax" is the angle in seconds of arc which the Earth's equatorial semi- diameter, or radius, subtends at the centre of the Sun, when the Earth is at its mean distance. As the equatorial diameter of the Earth is about 7926 THE ARITHMETIC OF ASTRONOMY. 283 miles, its radius will be 3963 miles. Hence we have the following simple proportion for finding the Sun's distance 8'8 : 206,265 : : 3963 : Sun's distance whence Sun's distance = 2 6>2 g 5 8 x3963 = 92,889,56; miles. To find the Sun's diameter, we know from observ- ation that, at the Sun's mean distance, the diameter measures 31' 3*6", or I923'6" of arc. Hence we have the proportion 206,265 : 1923*6 : : 92,889,567 : Sun's diameter. Hence Sun's diameter = 9 -^M31 6 = 8662 miles. To find what the Sun's density (or specific gravity) is, with reference to that of the Earth, we know from the principles of Geometry that the volumes of spheres are proportional to the cubes of their diameters. Taking the Sun's diameter in round numbers at 866,000 miles, and that of the Earth at 7912 miles, we have the volumes as (866,000) 3 : (79 1 2) 3 Working this, out, we find that the Sun's volume is equal to 1,311,400 times that of the Earth. But it has been found from mechanical principles that the 284 THE WORLDS OF SPACE. Sun's mass is only 327,214 times the mass of the Earth. Hence its density must be 327 ' 214 -^0-2495 1,311,400 that of the Earth being represented by I . If we take the Earth's specific gravity at 5 '6, that of the Sun will be 5-6 x 0-2495 = 1*3972 (water = i)= 1*40 nearly. The latest determination of the speed at which light travels through space gives a velocity of 186,337 miles a second. Hence to find the time taken by light to reach us from the Sun, we have merely to divide the Sun's distance by this velocity, or Time in seconds = ^^ = 498 seconds AOU) OO / or 8 minutes, 18 seconds. When we speak of the "parallax" of a fixed star, the word " parallax " has not exactly the same mean- ing as in the case of the Sun's distance. Here our base line of measurement is not the Earth's radius, but the radius of the Earth's orbit round the Sun, or the Sun's mean distance from the Earth. The parallax of a fixed star is then the angle which the radius of the Earth's orbit subtends at the star. The angle which is actually measured is of course the angle subtended by the diameter of the Earth's orbit, and is double the star's annual parallax. Hence for a THE ARITHMETIC OF ASTRONOMY. 285 star with an annual parallax of i" of arc, the distance from the Earth would be 92,889,567 x 206,265 = 19,159,866,537,255 miles, or about 19 billions of miles. There is, however, no star as far as we know at present with a parallax so large as i' of arc. The parallax of the nearest Alpha Centauri is about three-fourths of a second. Therefore to obtain its distance, we must divide the above result by f, or multiply by -J. This gives 25,546,488,716,340 miles, or about 25^ billions of miles for the distance of Alpha Centauri. To find the time taken by light to travel from this star to the Earth, we must divide the above number by 186,337, as before. This gives I37>O98,3I5, or 4*344 years, or about 4 years and 4 months. This calculation leads us to a simple expression for rinding the light journey from any star in years when its parallax is known. For we see that the distance for a parallax of i" had to be divided by the measured parallax, and again by the velocity of light, and by the number of seconds in a year, which is 365-^ x 24 x 60 x 60 = 31,557,600. 286 THE WORLDS OF SPACE. Hence if we call the parallax of the star /, we have , 1 9, i '59,866,? 3 7, 25 5 3*258 Light journey in years = _^^3^5j_ = ^_ For a parallax of one-tenth of a second the light journey would therefore be 32-58 years. This is about the parallax ascribed to the star 1830 Groom- bridge, which has a proper motion of about 7" of arc per annum. Let us see what this velocity will give. With a parallax of one-tenth of a second, the dis- tance travelled in one year will be 7" divided by -j 3 ^, or 70 times the Sun's distance from the Earth. Hence the velocity will be =-6 miles a second. Arcturus, with a proper motion of 2*26" of arc annually, has a parallax, according to Glasenapp, of only 0*018 of a second. This gives the enormous velocity of 368 miles per second. In general, the velocity may be simply obtained by multiplying the proper motion by 2-943, and dividing by the measured parallax. The velocities found in this way represent the star's motion on the background of the sky, or at right angles to the line of sight. There may also be and generally is motion in the line of sight, to or from the eye, but this can only be ascertained by the spectroscope. It may interest some readers to know how the THE ARITHMETIC OF ASTRONOMY. 287 mass of a binary or revolving double star is found in terms of the Sun's mass. In my Planetary and Stellar Studies I have given a formula for calculating the combined mass of the components of a binary star when the elements of the orbit and the star's parallax are known. The computation is a very simple one. Let us take the case of the binary star 70 Ophiuchi, for which I have computed an orbit, and find a period of 87*84 years (a period which has been confirmed by Burnham), with a mean distance between the components of 4-5" of arc. Now Kruger found a parallax of 0*162 of a second, and hence Mean distance between components = ^A~ 2 7'77 times the Sun's distance from the Earth, and Sum of masses of components = |||2| 2 = 2 77 times the mass of the Sun. The rule expressed in words is Divide the mean distance between the components by the parallax. Cube the quotient, and divide by the square of the period in years, and the result will give the sum of the masses of the components in terms of the Sun's mass, taken as unity. In astronomical works we find the force of gravity on the Sun and planets stated in terms of gravity at the surface of the Earth. If we know the relative masses of the two bodies, this is easily calculated. Let us take the case of the Sun and Earth. The 288 THE WORLDS OF SPACE. mass of the Sun is, as stated above, 327,214, if that of the Earth be taken as i. Now it is a well-known principle that the mass of a sphere acts on a body at its surface as if the whole mass of the sphere were concentrated at the centre of the sphere. Its action will also be inversely as the square of the distance from the centre, or as the square of the radius. Hence we have Force of gravity at Sun's surface = 327 = 27-41. Hence a body which would weigh i pound at the Earth's equator, would weigh 27-41 Ibs. if transferred to the surface of the Sun. The force of gravity at the surface of all the planets and satellites may be calculated in a similar way. XXVIII. THE MYTHOLOGY OF THE STARS AND PLANETS. MANY of the names of the constellation figures and those of all the larger planets are evidently derived from the ancient mythology of Greece and Rome. A short account of the origin of these names may prove of interest to the reader. We will first take the constellations in alphabetical order. Andromeda (the Chained Lady) was the daughter of Cepheus, King of Ethiopia, and his wife Cassiopeia. To assuage the wrath of Neptune, who instigated by the jealousy of the sea-nymphs for the vaunted beauty of Cassiopeia had sent g. great sea monster to ravage the whole country of Ethiopia, Andromeda was chained to a rock on the sea-shore, and was saved from destruction by Perseus, who, returning from the slaughter of the Gorgons with the head of Medusa, turned the monster into stone, and married Andromeda. Aquarius (the Water-Bearer) is one of the twelve signs of the Zodiac, and has been represented from ancient times as a man pouring water from a vase. U 2QO THE WORLDS OF SPACE. Aquila (the Eagle, Flying Vulture, or Flying Grype) formed one of the asterisms of Hipparchus. Its brightest star, Altair, derives its name from the Arabic word el-tatr, the flying eagle. Argo (the ship Argo). This large southern con- stellation formed one of the original 48 constellations. It represents the ship which conveyed the expedition in search of the Golden Fleece. Its brightest star, Canopus, is second only to Sirius in brilliancy, but does not rise above the English horizon. Aries (the Ram) represents the Golden Fleece which the Argonautic expedition went in search of. Some, however, have attempted to identify it with the ram sacrificed by Abraham on Mount Moriah ; but this seems very fanciful. Aries forms one of the twelve signs of the Zodiac. Auriga (the Charioteer or Waggoner) formed one of the original 48 constellations, and is thought by some to be the Horus of the Egyptians, but this seems doubtful. The brightest star, Capella, with e, C and r\ form the goat Amalthea, Jupiter's nurse. Capella has been called "The Shepherd's Star," a term which has also been applied to the planet Venus. Bootes (the Herdsman). According to Grecian mythology, Bootes was the son of Jupiter and Callista. He was a great hunter, and one day while on the chase he met a bear, which proved to be his mother transformed into this shape by Juno ! When THE MYTHOLOGY OF THE STARS AND PLANETS. 2QI about to kill the bear, Jupiter intervened and trans- ferred them both to the sky. Arcturus, the brightest star of the constellation, was well known to the ancient mariners, and its name appears to be derived from two words signifying "the bear's tail." It seems to have been the first star observed in daylight with the telescope by Morin in 1635. The Arcturus mentioned in Job probably refers to the Great Bear. Cancer (the Crab). This is one of the original 48 constellations, and is said to represent a crab which was raised to the skies because it pinched the toes of Hercules in the Lerncean marsh ! In some old star maps it is represented as a crayfish or lobster. Cancer forms one of the signs of the Zodiac. Cants Major (the Great Dog) formed one of the ancient 48 constellations. Its brightest star, Sirius, derives its name from the Greek word Sei'ptoy, and it was worshipped by the Egyptians under the name of Sothis (Horus), Anubis, and Thoth. Some consider it to be the Mazzaroth of Job. Sirius was also sup- posed to be Orion's hound, and may perhaps be identical with the Cerberus of the Greeks. Cants Minor (the Little Dog) also formed one of the original 48 constellations, and was called UpoKvcov, the precursor, because it appeared in the morning sky before Sirius. The name survives in Procyon, its brightest star. Capricornus (the Sea Goat). This is also one of the old 48 constellations, and was supposed to repre- THE WORLDS OF SPACE. sent the plunge of the god Pan into the Nile. Pan is said to have been the son of Mercury and Penelope, and was represented as half man and half goat. Capricornus forms one of the signs of the Zodiac. Cassiopeia (the Lady in the Chair) was the wife of Cepheus, to whom the adjoining constellation is dedicated. She is represented as bound to a throne or chair formed by the five well-known stars arranged like a W. Centaurus (the Centaur) is a southern constellation, and probably derives its name from the fabled cen- taurs of mythology, who were half men and half horses. Cepheus (the Monarch), King of Ethiopia. He was the husband of Cassiopeia, and father of Andromeda. Coma Berenices (Berenice's Hair) is stated by Conon to represent the tresses of a lady placed in the heavens as a reward for a lock of hair which on account of a victory won by her husband she had dedicated to Venus. Corona Borealis. According to Plutarch this repre- sents a crown given by the god Bacchus to his wife Ariadne. Corvus (the Crow), supposed by some to represent the raven sent out by Noah from the ark. Cygnus (the Swan). This fine constellation, one of the glories of the Northern sky, formed one of the original 48 asterisms. Its principal stars form a fairly regular cross, and hence it was called Christi THE MYTHOLOGY OF THE STARS AND PLANETS. 293 crux by Schickard and others in the eighteenth century. Delphinus (the Dolphin). One of the original 48 constellations. According to some of the ancient writers it represents Apollo bringing Castalius from Crete. Others suppose it to have been the dolphin which carried Arion. Novidius, however, saw in it the fish which swallowed Jonah ! Draco (the Dragon), which guarded the golden apples in the garden of the Hesperides, and was killed by Hercules. Some, however, think it typical of the serpent which tempted Eve ! Eridanus (the River). Some suppose this repre- sents the Nile, others say the Po, or the Granicus, or the Euphrates. It has also been called the River of Orion. Eridanus lies to the south-west of the highest point attained by the Sun on the longest day ; and this may have some connection with the fable of Phaeton, who, after driving the chariot of the Sun, was struck dead by Jupiter, and fell into the river Eridanus. The brightest star of this constellation is Achernar (the end of the river). It does not rise above the English horizon. Gemini (the twins), Castor and Pollux, sons of Jupiter and Leda. They accompanied Jason in his expedition in search of the Golden Fleece, and were conspicuous for their bravery. Jupiter made Pollux immortal, but not Castor. When Castor was slain by Lynceus, Pollux shared his immortality with 2Q4 THE WORLDS OF SPACE. his brother, and they lived and died alternately. They were finally placed in the sky, and changed into the two bright stars which now bear their names. Gemini forms one of the signs of the Zodiac. Hercules, the great hero of mythology, was supposed to be the son of Jupiter by Alcmene, wife of Amphy- trion. Hercules possessed great strength and courage, and his twelve labours are familiar to students of the classics. One of these labours was the destruction of the dragon which guarded the garden of the Hes- perides (see Draco). Hercules formed one of the 48 ancient constellations. It is of considerable extent, but includes no star brighter than the third magnitude. Hydra. Supposed to represent the Lernsean ser- pent. It was said to have had 50 heads, and that when one was cut off many others were produced in its place. It was finally killed by Hercules. Z^(the Lion). Supposed to represent the Nemean lion killed by Hercules. Stower, however, in 1386, thought it represented one of Daniel's lions ; and it was considered by Schickard as typical of " the Lion of the tribe of Judah." Its brightest star, Regulus, has been called the royal star. Lepus (the Hare) was one of the original 48 con- stellations. The right foot of Orion rests on Lepus. Libra (the balance). One of the signs of the Zodiac. According to some writers this constellation was formed by the Alexandrian astronomers ; but others say it was raised to the memory of Julius Caesar. It THE MYTHOLOGY OF THE STARS AND PLANETS. 295 seems to have been known to Ovid, Pliny, Virgil, Vitruvius, and other ancient writers. Lyra. Supposed to represent the lyre of Orpheus, a famous Grecian poet who accompanied the Argo- nautic expedition in search of the Golden Fleece. Novidius, however, saw in it the harp of David, and Schiller supposed it to represent the manger of Bethlehem ! It formed one of the original 48 con- stellations. Its brightest star, Vega, is one of the most brilliant in the heavens. Ophiuchus (the Serpent Bearer) is one of the old 48 constellations, and is supposed by some to represent one of Hercules' victories. It has been called Serpentarius. Orion (the Heavenly Hunter) was said to have been the son of the god Poseidon and Euryale, daughter of Minos. According to some, however, his mother was Alcyone, one of the Pleiades. Orion was said to have been an attendant on the goddess Diana (or Artemis). Having one day insulted his mistress, he was either killed by a scorpion, or shot by the goddess herself, and was then placed among the stars. Orion was worshipped in Egypt under the name of Osiris. He has also been identified with the Nimrod of Scripture (Genesis x. 8, 9). Homer speaks of Orion marrying Aurora, goddess of the Dawn. The fable of his having been killed by a scorpion probably refers to the disappearance of Orion in the west when the constellation of the 296 THE WORLDS OF SPACE. Scorpion appears above the eastern horizon. Some suppose that Adonis (Tammaz) was identical with Orion. Merodach, the god of Babylon, seems to have been Orion under another form. Orion is mentioned by Job, Ezekiel, and Amos. Pegasus (the Winged Horse) is supposed to represent Apollo's horse, but by others the horse of Nimrod. Perseus (the rescuer of Andromeda) was the son of Jupiter and Danae. This constellation formed one of the 48 original asterisms, but according to Schickard's innovations it represents David with the head of Goliath ! Pices (the Fishes). This also formed one of the old 48 constellations. It is supposed to represent two fishes with their tails tied together with a cord. It is one of the signs of the Zodiac. Sagitta (the Arrow). Small as this constellation is, it formed one of the ancient 48. It lies between Aquila and the head of Cygnus. Sagittarius (the Archer) is usually represented as a figure like the ancient centaurs, who were fabulous creatures, half man and half horse. It has, however, sometimes been drawn as a satyr. It is one of the twelve signs of the Zodiac. Scorpio (the Scorpion). This was one of the original 48 constellations, and may perhaps represent the scorpion which killed Hercules. It forms one of the signs of the Zodiac. Its brightest star, Antares, THE MYTHOLOGY OF THE STARS AND PLANETS. 297 is so named from the Greek words meaning " redder than Mars." Serpens (the Serpent). This also formed one of the old 48 asterisms. It was called by the Greeks O a value which actual measurements show to be too large. Clairaut sub- sequently considered the subject on the hypothesis that the Earth's density increases towards the centre, and obtained much better results. Observations have shown that the weight of a body at the Earth's 310 THE WORLDS OF SPACE. poles would exceed the weight of the same body at the equator by about T i T th part of the whole weight ; that is, a body weighing 187 Ibs. at the equator would weigh 1 88 Ibs. at the pole. Clairaut showed by mathematical analysis that the sum of the fractions representing the ellipticity and that representing the increase of gravity at the poles will be equal to f of the ratio between the centrifugal force at the equator and the force of gravity. Now, knowing the dimensions of the Earth and the time of rotation on its axis, we can calculate the amount of the centrifugal force at the equator, and this comes out -g^-g-. Hence we have by Clairaut's theorem the equation (e = Earth's ellipticity) ST ~ 2~"ir X -- "ir - Whence e a value which agrees closely with the results of actual measurement. It may be here explained that the fraction repre- senting the ellipticity or flattening at the poles is found by subtracting the polar diameter from the equatorial, and dividing the difference by the equa- torial diameter. The difference of the diameters in the terrestrial spheroid is about 26^ miles. Several measurements of arcs were made in the eighteenth century by Maupertius, Clairaut, Le Mon- nier, and others, and the results showed that the Earth is an oblate spheroid. THE FIGURE OF THE EARTH. 311 Combining all the available data, the dimensions of the terrestrial spheroid were computed by Bessel and Clarke. Bessel in 1841 found the equatorial diameter of the Earth to be 7925-52 miles, and the polar diameter 7899*03 miles, indicating a compression or ellipticity of -^^. In 1859 General de Schubert suggested that the discrepancies in the data might be reconciled by supposing the figure of the Earth to be not an exact spheroid of revolution, but a figure known as an ellipsoid, that is a solid in which the plane passing through the centre at right angles to the shorter axis is not a circle, but an ellipse. On this hypothesis the Earth would have three axes of different lengths at right angles to each other. The lengths of these three axes were computed by General de Schubert, and afterwards by General Clarke. The great physical improbability of the Earth having a figure of this kind has, however, led mathematicians to reject the ellipsoidal hypothesis and to attribute the apparent deviations from the true spheroidal figure to the effects caused by local attraction of mountains, etc. on the plumb-line of the instruments used in determining the latitudes. In 1880 General Clarke abandoned the ellipsoidal hypothesis and reverted to the theory of an oblate spheroid. Using data derived from arcs measured in Russia, India, the Cape of Good Hope, Peru, and 312 THE WORLDS OF SPACE. the Anglo-French arc, he arrived at the following results- Feet. Miles. Earth's equatorial diameter = 41,852,404 = 7926^59 polar = 41,709,790 = 7899-58 From a discussion of all the available data Professor Harkness found in 1891 the following results Equatorial diameter = 7926*248 + 0-156 miles Polar = 7899^44 + 0-124 Flattenin g = 300-205^2-964 Pendulum experiments since 1830 give value for the flattening ranging from -gw-Tir to inri-Te-> tne mean of 14 determinations being ^W-TTI an< ^ Professor Harkness concludes from these experiments that " either the flattening must lie between i : 296 and i : 300, or the distribution of density within the Earth cannot be represented by any function which in- creases continuously from the surface to the centre." Considering the results derived from all the methods of calculation, including those from precession and nutation, and the perturbations of the Moon, he con- siders that "the facts thus far adduced scarcely warrant any conclusion more definite than that the flattening lies between i : 290 and i : 300, but with some further evidence which tends in the direction of the smaller limit." THE FIGURE OF THE EARTH. 313 The comparatively close agreement, however, between the results found by different methods shows clearly that the spheroidal theory is correct. We may therefore conclude that the figure of the Earth is practically an oblate spheroid of revolution with a flattening or ellipticity of about -3%$. XXX. THE VARIATION OF LATITUDE. To most people, even those familiar with the facts of astronomy, the latitude of any place on the Earth's surface would probably be considered as an invari- able quantity, constant at all periods of the year, and the same for every year. To mathematicians ac- customed to discuss and calculate the rotation of a rigid body as the Earth is supposed to be the constancy in position of the Earth's axis of rotation would be taken for granted. And, indeed, for all practical purposes it might be justly concluded that the latitude does not vary. Ordinary observations would show that the celestial pole maintains a constant elevation above the horizon, this elevation, or altitude, being equal to the latitude of the place of observation. Careful measures, how- ever, made during a long series of years at various astronomical observatories have shown that the lati- tude is really subject to a small variation, which be- THE VARIATION OF LATITUDE. 315 comes of considerable importance when the modern refinement of astronomical calculations is considered. In a remarkable series of papers, published in 1892 and 1893, by the well-known American astronomer Mr. S. C. Chandler, he gives the results of his researches on this interesting subject. His first suspicion of a periodical change of latitude seems to have been aroused some eight years ago, when observations with his newly-invented almucantor appeared to indicate a change of latitude. The correctness of this result was apparently so improb- able that Mr. Chandler merely published his observa- tions, and waited paitently for further evidence in support of this startling discovery. In the year 1888 Dr. Kustner found similar irregularities in his observ- ations of latitude, and these were confirmed by observ- ations made at several observatories on the continent of Europe. Encouraged by these results, Mr. Chandler then resolved to rigorously attack the problem, and adopt- ing a provisional period of 400 days for the supposed variation in latitude, he analyzed a long series of observations made at observatories between the years 1726 and 1887. His investigations which involved the manipulation of an enormous number of figures showed evidence of variation of the latitude, due, he considers, to a rotation of the Earth's axis from west to east in a period of about 427 days, and in a circle 316 THE WORLDS OF SPACE. having a radius of 30 feet measured on the Earth's surface. In other words, his results indicate that the Earth's axis of rotation does not coincide with the axis of inertia. From a further investigation this preliminary result was somewhat modified. The observed variation was found to be the result of two periodical variations, one with a period of about 427 days, the other having a period of one year. These two fluctuations some- times combine their effects, and at others tend to neutralize each other. The maximum effect of the annual variation occurs about the time of the autumnal equinox (Sept. 23), and the minimum about the vernal equinox. The maximum combined effect of the two variations may amount to about 0*6 of a second of arc. From observations of circumpolar stars made at the Greenwich Observatory in the years 1851 to 1891, Messrs. Thackeray and Turner find results which tend to confirm those of Mr. Chandler, and afford strong evidence in favour of his theory. Concerted observations made at Berlin, Potsdam, and Prague in the years 1889 1891 fully confirm the variation in latitude. They indicate a fluctuation of about half a second of arc annually, and the observ- ations are wonderfully accordant. Mr. Chandler's conclusions are supported by such eminent astronomers as Professor Simon Newcomb THE VARIATION OF LATITUDE. 317 and Dr. 13. A. Gould, at least as far as the present period of variation is concerned. Mr. Chandler's investigations show that, although the period is now 427 days, it was less than one year about the year 1770. Professor Newcomb, however, maintains that the length of the period must be invariable, and he refers to the elasticity of the Earth as a possible physical explanation of the latitude variation. It has been suggested that the variation in latitude may, perhaps, be due to a meteorological effect ; that is, that it may possibly be caused by periodical changes in the arrangement of the atmospherical strata, which would of course affect the refraction, and hence the astronomical observations. But if this were so, we should expect the variation to show an annual period, and not one of some 14 months. Dr. Herz suggests a theory which connects the variation with an electrified state of the Sun. He supposes the Earth not to be a perfect conductor, and the air an imperfect non-conductor, and he shows that the direction of gravity (or the direction of the plumb- line) will be subject to an apparent change, which would affect the observations of latitude. Whatever the physical cause may be of this un- expected variation in the latitude, it seems demon- strated beyond a doubt that a small periodical change does take place. At first sight it may seem that the observed varia- 318 THE WORLDS OF SPACE. tion is so small that it might be neglected for all practical purposes. For geodetical operations this may be true, but for the refined requirements of astronomy, such a variation in latitude formerly assumed to be invariable vitiates the determination of various astronomical constants, the computation of which was based on the assumption of a constant zenith. Some of the constants thus affected are, the equinoctical points, the obliquity of the Ecliptic, the constant of aberration, the measures of parallax, the system of right ascensions and declinations, and possibly, to a small extent, the constants of precession, nutation, and refraction. The constant of aberration of the stars (an apparent effect due to the Earth's orbit round the Sun, com- bined with the progressive motion of light) was for many years considered by astronomers as known with great accuracy, the only doubt being as to the value of the second place of decimals (the value ranging between 20*447" an d 20*492" as determined by different computers). Owing to the discovery of the latitude variation, however, Mr. Chandler thinks that even the value of the first decimal is now called in question. Making due allowance for the variation of latitude, he deduces from Peters' observations of the Pole star in the years 1842-44 a value of 20*510" for this important astronomical constant. He also thinks that " hitherto unexplained discordances among the THE VARIATION OF LATITUDE. 319 various determinations " of the equinoctial point, or zero-point of right ascensions, may be reconciled by making due allowance for the variation of latitude. Combining the value given above for the constant of aberration with Newcomb's determination of the velocity of light, he finds the solar parallax to be XXXI. THUNDERSTORMS AND AURORAS. IN some very interesting papers recently read before the Academy. of Sciences, Rochester (U.S.A.), Dr. M. A. Veeder gives the results of his researches on Thunderstorms, Auroras, and the Zodiacal Light. He finds that Auroras tend to recur at intervals of a synodical rotation of the Sun on its axis, which is about 27 days, 6 hours, and 40 minutes. The Sun really rotates in a period of about 25 J days, but owing to the Earth's orbital motion round the Sun in the same direction as the Sun's rotation, this period is . apparently lengthened to about 2/J days. The Sun's axis of rotation is inclined to the plane of the Ecliptic the plane of the Earth's orbit at an angle of about 83, and consequently the Sun's Northern Hemisphere sometimes leans towards the Earth, and sometimes his Southern Hemisphere. Dr. Veeder finds that in general "Auroras and their attendant magnetic storms occur when spots or faculae, or both, are at the Sun's eastern limb, and THUNDERSTORMS AND AURORAS. 321 near the plane of the Earth's orbit/' He also finds that the recurrence of Auroras is not continuous, but that they occur more frequently near the equinoxes (March 21 and September 23), the phenomena almost disappearing near the solstices (June 21 and December 22). He has also discovered that thunder- storms show a tendency to take the place of Auroras when the latter fail to appear, and he concludes from this relation that both phenomena have their origin in electrical action between the Sun and the Earth. There is, however, a slight difference between the conditions favourable to the production of the pheno- mena. It appears that thunderstorms generally take place when the disturbed portions of the Sun are at the eastern limb and " at a distance from the plane of the Earth's orbit," while Auroras usually occur when spots and faculae are close to the same plane. Dr. Veeder finds that thunderstorms usually occur in the afternoon, with a "secondary maximum between midnight and morning." Thunderstorms are of such frequent occurrence in India that it has occurred to me to test Dr. Veeder's theory by the meteorological records of that country. With this view I have care- fully examined the records of thunderstorms as given in the Madras Meteorological Results for the years 1 86 1 1890, and I find a well-marked tendency to occurrence in the afternoon and early morning hours. They also show a tendency to occur with greater 322 -THE WORLDS OF SPACE. violence near the equinoxes than at other times. The following are the dates of severe thunderstorms during the thirty years of observation at Madras Observatory 1863 July 10 " heavy thunderstorm " 1871 Sept. 27 "heavy" 1873 Oct. 3 "heavy" 1876 April 30 "heavy" 1879 Oct. 8 " severe thunderstorm " 1880 Sept. 9 "heavy" 1 88 1 Aug. 6 "heavy" 1884 Oct. 16 "heavy" 1886 Oct. 31 "heavy" 1887 Sept. 29 "heavy" 1887 Oct. 9 " cyclone and thunderstorm " 1890 July 26 "heavy" Of these 1 2. great thunderstorms, nine occurred not far from the equinoxes. Of the three exceptions, two took place in July, about the beginning of the monsoon (or yearly rain season), when the atmospheric con- ditions are usually much disturbed. There seems, therefore, to be strong evidence in favour of Dr. Veeder's conclusions. In the year 1880, thunder- storms occurred on July 19 at 5j p.m., and on August 15 at 8 p.m., the interval being 27 days 2\ hours in close agreement with the period of the Sun's synod- ical revolution and again on Oct. 4 and 31, an interval of 27 days. Records of Auroras in the United States tend to confirm the truth of. Dr. Veeder's hypothesis. Thus, Auroras were visible in the year 1892 on January 5, THUNDERSTORMS AND AURORAS. 323 February 2, February 29, March 27, and April 23, or at intervals of 27 days, and " associated with reappear- ances at the Sun's eastern limb of an area south of the equator which has been much frequented by spots and faculae." Dr. Veeder advances a new view of the Zodiacal Light which has long been an enigma to astronomers. He considers that the light is an extension of the solar corona, and " is probably double, corresponding to the bifurcation seen during eclipses, each section overlying a sun-spot belt. In the spring months, the south pole of the Sun being turned towards the Earth, the southern section is seen edgeways and becomes invisible (like Saturn's rings), the northern section being at the same time opened out to its widest extent, and reflects far more light earthward than at any other time." The light is then seen after sunset. This state of things is reversed in the autumn, when the southern section becomes visible before sunrise. In winter and summer "the Earth occupies a position intermediate between these disc- like coronal extensions," and consequently the Zodi- acal Light is less conspicuous. Dr. Veeder is of opinion that the Zodiacal Light consists of ferrugi- nous particles, and that it is the medium by which magnetic phenomena are transmitted from the Sun to the Earth. If Dr. Veeder's theory is correct, we should expect to find some relation between the periodical development of sun-spots and the pheno- 324 THE WORLDS OF SPACE. mena of Auroras and magnetic storms, and such a connection has been actually shown by observation, the number of Auroras and magnetic disturbances being more frequent in the years of sun-spot maxima. XXXII. THE BAROMETRIC MEASUREMENT OF HEIGHTS. ] THERE are several methods of measuring the heights of mountains and other elevated portions of the Earth's surface above the sea-level. Of these may be mentioned the following: (i) By actual levelling with an engineer's spirit-level and graduated staff ; (2) by trigonometrical calculation based on the measurement of the angles of elevation observed at the extremities of a carefully-measured base line ; (3) by observing the temperature of the boiling-point of water ; and (4) by reading a barometer at the sea- level, and again at the top of the mountain or eleva- tion, the height of which is to be determined. The first of these methods is certainly the most accurate, but it involves a considerable amount of labour, and for very high mountains is sometimes impracticable. The second method is sufficiently accurate, if carefully carried out, and a nearly level plain is available for the measurement of a base line. The third method is not accurate enough to give reliable results. The fourth is the simplest and most 326 THE WORLDS OF SPACE. expeditious of all. It is especially useful for finding the difference of level between two points at consider- able distances apart, and would be sufficiently accurate if certain difficulties could be successfully surmounted. A consideration of this method, and the difficulties to be overcome, before its accuracy can be relied upon, may prove of interest to the general reader. The principle of the barometric method is as follows : The barometer measures the weight or pres- sure of the atmosphere. The column of mercury in an ordinary mercurial barometer is equal in weight to a column of air of the same diameter, and of a height equal to that of the Earth's atmosphere. The densest portion of the atmosphere is that close to the Earth's surface, and its density diminishes as we ascend. At the top of a mountain, therefore, the pressure of the atmosphere will balance a shorter column of mercury, and hence the mercury descends in the tube. From the difference in height of the mercury at the level of the sea, and on the top of the mountain, it is possible to calculate the height we have ascended, as will be shown further on. There are two forms of barometers, namely, the mercurial barometer and the aneroid. Of mercurial barometers, there are two forms, the " cistern " and the " syphon." The cistern form is the one generally used for scientific observations, and is the best for measuring heights. One of the most approved forms of cistern barometers known as " Fortin's barometer " THE BAROMETRIC MEASUREMENT OF HEIGHTS. 327 consists of a glass tube, closed at one end, and filled with mercury, the lower portion of which dips into another tube of larger diameter, which contains a reservoir of mercury forming the "cistern." The bottom of the cistern is formed of leather, and fitted with an adjusting screw below, for the purpose of adjusting the level of the mercury in the cistern to an ivory index point, which marks the zero of the graduated scale used for reading the height of the mercurial column. By means of this adjusting screw the mercury may also be raised so as to completely fill the cistern and tube, and thus adapt the instru- ment for travelling. We need not discuss here the manufacture of barometers, and the filling of the tube with mercury, an operation which must be performed carefully so as to exclude air from the tube. Suffice it to say, that the best method is to fill the tube gradually, and boil the mercury as we proceed by means of a spirit- lamp, in order to drive out all bubbles of air which may be contained in the mercury. The tube may be filled without boiling, but the resulting instrument will not be so accurate as one in which the mercury has been boiled. To determine the difference of elevation between two places with a mercurial barometer, several points must be attended to. In the first place, the temper- ature of the barometer and the temperature of the air must be noted at each station. As the mercury 328 THE WORLDS OF SPACE. in a barometer is affected by heat in the same way that a thermometer is the temperature at which the barometer is read must be observed. For this purpose a thermometer is usually attached to the barometer. The temperature should be read as accurately as possible, for an error of one degree Fahrenheit would make a difference of about three feet in the resulting altitude. The reading of the attached thermometer should be first noted, and then the height of the barometer. To do this, first bring the surface of the mercury in the cistern accurately to the index-point by means of the adjusting screw. Then tap the tube gently near the top of the column in order to get rid of the adhesion between the mercury and the glass. The height of the mercury may then be read by means of the attached scale and vernier. Sometimes the amount of aqueous vapour in the atmosphere is ascertained by another instrument. The above data being known for two stations, we substitute the values found in one of the barometric formulae, and thus obtain the height or difference of height required. Before the barometer readings can be used, they must be reduced to the same temperature usually 32 degrees Fahrenheit. Various formulae have been computed by eminent mathematicians and physicists for calculating the difference of height between two points. These formulae depend on certain assumptions, which, how- ever, cannot be considered as rigidly true. The most THE BAROMETRIC MEASUREMENT OF HEIGHTS. 329 important of these assumptions is, that the atmosphere may be considered to be in a state of statical equili- brium. But owing to the changes constantly taking place, due to differences of temperature, humidity, winds, etc., this assumption cannot be considered as correct. The result will therefore be only an approxi- mation to the truth. Assuming, however, a statical equilibrium of the atmosphere, a formula can be easily deduced from known principles. For this purpose we must first ascertain the weight of a cubic inch of air and a cubic inch of mercury at a certain temperature and pressure, and in a given latitude, say 45. Then, by Boyle and Marriotte's law con- necting the weight of a gas and the pressure, a formula can be found for determining the height required. There are several elaborate formulae used for the purpose. These include terms for altitude, latitude, temperature, and humidity. A correction for altitude is theoretically necessary, owing to the diminution in the force of gravity and, therefore, a decrease in the weight of bodies with increased distance from the centre of the Earth, but this correction is comparatively very small, and may, for all practical purposes, be neglected. For the same reason a correction for latitude is mathematically required, owing to the spheroidal figure of the Earth ; but this, too, is very small, and may be safely neglected. The correction for temperature of the air is, however, very important. This term is easily computed. It is obtained for 33O THE WORLDS OF SPACE. the Fahrenheit scale by deducting 64 from the sum of the observed temperatures at the upper and lower stations, dividing the difference by 900, and adding unity to the result. A correction for the humidity of the air is also necessary ; but it is doubtful whether it is desirable to complicate the formula by a correc- tion for atmospheric moisture, the laws of which are so imperfectly understood. In all the barometric formulae which have been proposed, the first term is constant and common to all. It is known as the " barometric co-efficient," and is 5744^, where mis the "weight of a cubic inch of mercury at the sea-level in latitude 45 at 30 Fahrenheit, when the barometer reads 29*92 inches," and a the weight of a cubic inch of dry air under the same conditions of latitude, temperature, and pressure. Various values of this constant have been found, depending on the values assumed for m and a. Arago and Biot found ^=10,467. This makes the "barometric co-efficient" 60,122-4 feet Raymond's value, namely 60,158*6 feet, was found by comparing the values given by the formulae with the results of actual levelling with a spirit-level. His observations were, however, few in number, and although his co-efficient is often used, it is probably the least accurate of all the determinations. In Laplace's formula Raymond's constant is used. Babinet used the constant 60,334, and in Baily's formula the constant is 60,384, which is the value found by THE BAROMETRIC MEASUREMENT OF HEIGHTS. 331 Regnault, and is probably the most accurate of all. Sometimes the co-efficient in the formula is given as 10,000 fathoms > which is roughly correct. We will now consider the errors underlying the barometric measurement of heights, and which render the method inapplicable in cases where great accuracy is required. The most important of these sources of error is probably that due to what is called the " barometric gradient," a term frequently used in meteorological reports. Taking three points at which the barometric pressure is the same ; if the atmosphere was in a state of statical equilibrium these points would lie on the same level plane. But usually this plane is not level but inclined, and the inclination of the plane is termed the " barometric gradient." For a number of points the surface on which they lie would not be a plane at all, but an undulating surface. These surfaces for different heights are never parallel, and frequently slope in opposite directions. Allowance cannot be fully made for this disturbing cause, but the error can to some extent be eliminated by making a number of simultaneous observations at the two stations and taking a mean of the results. Another cause of error is due to variations in the temperature of the air. It is generally assumed that the mean temperature of the column of air between two^stations, one vertically over the other, is the mean of the temperatures at the upper and lower 332 THE WORLDS OF SPACE. stations ; but this is not always the case. The error may be partly eliminated by making observations at intermediate stations, but cannot be entirely overcome. High winds also cause a variation in the height of the barometer. In addition to the errors mentioned, there are, of course, errors of observation and instrumental errors. The former may be caused by imperfect adjustment of the zero-point and erroneous reading of the mercury on the scale. These errors are, however, usually small, and may with care be neglected. The instrumental errors are chiefly due to imperfect graduation of the scales of the barometer and attached thermometer, impurity of the mercury, and to air in the tube. These errors may be corrected by com- parison with a standard instrument. The form of barometer known as the aneroid is also frequently used for the determination of heights, a graduated scale being added for that purpose. This scale is graduated by means of one of the barometric formulae already referred to. The aneroid barometer usually consists of a metallic box from which the air has been exhausted, and differences of atmospheric pressure are recorded by a system of levers which act on an index-hand which marks the reading on a graduated scale. In some forms of aneroid the box is not completely exhausted of air, and these are called " compensated aneroids," but the name is misleading, some of these instruments being THE BAROMETRIC MEASUREMENT OF HEIGHTS. 333 more sensitive to changes of temperature than those not compensated. The aneroid is a very handy instrument and easily used, but for the purpose of measuring heights it is much inferior to the mercurial barometer. In some aneroids the altitude scale is fixed at a certain reading, say 30 or 31 inches, and in others it is movable, and can be adjusted to any reading required. The latter seems the most con- venient plan. In either case it is clear that absolute elevations above the sea-level cannot be determined with this instrument with any approach to accuracy, as there is no way of making the necessary corrections for variations of pressure, temperature, etc. The aneroid barometer should, therefore, be used only for finding differences of elevation, and for this purpose it will give fairly good approximate results in cases where extreme accuracy is not required. To show the degree of accuracy attainable by the barometric method, two examples may be cited. From readings of a mercurial barometer at the summit of Mont Blanc and at the Geneva Observatory made by Messrs. Bravais and Martins in the year 1844, the height of the mountain above the level of the sea was computed to be 4815*9 metres, or 15,800*44 feet. Corabeuf found by trigonometrical measurement a height of 15,783 feet, or 17*44 feet less than that indicated by the barometer. The height of Mount Washington in the United States was found by a spirit-level to be 6293 feet 334 THE WORLDS OF SPACE. above sea-level, while the barometric method gave 6291*7, a close approximation. In some cases, how- ever, much larger differences have been found, and the good agreement quoted above may perhaps be considered as accidental. XXXIII. THE OBSERVATORY ON MONT BLANC. THE bold idea of founding a Meteorological Observatory on the summit of the highest mountain in Europe is due to the famous French astronomer, M. J. Janssen, who made an ascent of Mont Blanc in the year 1890, to enable him to judge of the possi- bility of such an undertaking. Having come to the conclusion that the project was feasible, M. Janssen made an appeal for funds, and numerous generous donations were subscribed for the purpose. A com- mittee was then formed, of which the President of the French Republic was an honorary member, and which included the names of Prince Roland Buona- parte, Baron Alphonse de Rothschild, M. Bischoffseim and others. The first thing to be considered was, of course, the possibility of obtaining the necessary foundations for the proposed building. The summit of Mont Blanc consists of a thick cap of ice or frozen snow, of unknown depth ; and to ascertain, if possible, the thickness of this permanent crust, borings were made 336 THE WORLDS OF SPACE. near the summit, with the hope of reaching the solid rock. These shafts, which were about 1 5 feet long, and about 27 feet vertically below the actual summit of the mountain, were driven horizontally at an angle of 45, but failed to meet with any rock. The idea then occurred to M. Janssen that it might perhaps be possible to found the building on the ice itself, which observations have shown to be subject to merely small and periodical changes. To ascertain whether the ice-cap would be sufficiently strong to bear the weight of a building, experiments were made in France. A mound of snow of the proposed height of the first floor was piled up during the winter at the Meudon Observatory, in such a way as to have the same density as that which covers the summit of Mont Blanc, at a depth of 3 to 6 feet below the surface. This density is, according to the measures of Lieu- tenant Dunod, about half that of water. The summit of this mound having been levelled off, it was loaded with discs of lead, each about 14 inches in diameter, and weighing about 66 Ibs. When a column of 12 discs had been raised, weighing about 792 Ibs., they were removed, and the depth of the impression made in the snow was measured. This amounted to less than one-third of an inch, a compression unexpectedly small, and proved that the ice-cap on the summit of Mont Blanc was abundantly strong for the support of a building about 33 feet long by i6J feet wide, which it was proposed to erect. THE OBSERVATORY ON MONT BLANC. 337 The stability of the foundation being thus demon- strated, it was necessary to consider the form of building most suitable to resist the violent storms which are of frequent occurrence on the summit of the mountain. The form of a truncated pyramid was selected, with the lower storey of the building buried in the snow. These preliminary investigations having been completed, it was decided to proceed with the construction of the building, and a suitable edifice was designed of the above form by an eminent architect, M. Vaudremer. The building is of two storeys, with a terrace and balcony, the lower storey being entirely below the surface of the snow. A spiral staircase runs the whole height of the building, connecting both storeys with the terrace, and rising above it several feet for the support of a small plat- form for meteorological observations. The walls are double, to protect the observers from the cold, and the windows and openings are provided with outside shutters which fit very closely. The building is furnished with heating apparatus and all the necessary accessories for life at this high altitude. The Observatory was then constructed and con- veyed in pieces to Chamounix. About three- fourths of the materials were at the end of 1892 transported to the summit of the Grands Mulcts, a height of about 9800 feet, and about a quarter to the Rochers Rouges (about 14,800 feet). The Observatory was practically completed towards the close of the year 1893. The 338 THE WORLDS OF SPACE. Observatory will be of an international character, and all observers who wish to visit it will be cordially welcomed. M. Janssen made another ascent of Mont Blanc in September 1893, in order to inspect the Observatory, and to make some observations on the solar spectrum with reference to the disputed question of the exist- ence of oxygen in the Sun. The observations were made with a grating spectroscope by Rowland on the B group of lines of the solar spectrum. This group includes a series of double lines, 13 or 14 doublets being visible at the surface of the sea. M. Janssen found that at Chamounix the thirteenth doublet was difficult to see. At the Grands Mulcts, only the tenth or eleventh pair could be seen. Finally, at the summit of Mont Blanc, M. Janssen could hardly see beyond the eighth pair, and from these observations he con- cludes that all the lines in the B group would dis- appear from the solar spectrum near the limits of the Earth's atmosphere. The conclusion then is that these lines, which are due to oxygen, are produced solely by the Earth's atmosphere, and that oxygen does not exist in the Sun. THE END. SOCIETY IN CHINA. By ROBERT K. DOUGLAS, KEEPER OF THE ORIENTAL BOOKS AND MANUSCRIPTS IN THE BRITISH MUSEUM; PROFESSOR OF CHINESE AT KING'S COLLEGE. An Account of the Every-day Life of the Chinese People Social^ Political^ Religious. With numerous Illustrations, reproduced in Collotype from Original Drawings by Chinese Artists, and from Photographs. 432 pp. Demy 8vo, Mandarin cloth extra. Price i6s. POEMS. By MOSSE MACDONALD. Fcap 4to, buckram, gilt top, untouched edges. Price 6s. nett. SEERS AND SINGERS. A STUDY OF FIVE ENGLISH POETS. By ARTHUR D. INNES. Fcap 8vo, antique cloth extra, gilt top, untouched edges. Price 5.? LONDON : A. D. INNES & Co., 31 & 32 BEDFORD ST., STRAND. BOOKS FOR PRESENTS AND PRIZES. Studies and Stories. By Mrs. MOLESWORTH. With a Frontis- piece by WALTER CRANE. Crown 8vo, $s. 6d. " Mrs. Molesworth's writings are always charming, and this new volume is one of her very best." Athenaum, The Dawn of History. By C. F. KEARY, F.S.A. Crown 8vo, 3.?. 6J. Church Lore Gleanings. By T. F. THISTLETON DYER. Illustrated. Large crown 8vo, $s. "One of the pleasant and painstaking books which are coming out now . . . gracefully and unpretentiously written . . . the selection is well chosen." Athenceum. Nature and Woodcraft. By JOHN WATSON. Illustrated by G. E. LODGE. Crown 8vo, $s. 6d. " It is long since we have met with a more delightful book on rural subjects than ' Nature and Woodcraft.' " Times. " This is the best written and most valuable of Mr. Watson's books." Academy. Biographies of Good Women. By CHARLOTTE M. YONGE. First Series, fcp. 8vo, well bound in cloth, 472 pages, 3^. 6d. Second Series, fcp. 8vo, well bound in cloth, 580 pages, 3^. 6d. The Soul's Hour Glass. Translated from the Horologium of Drexelius. Edited by the Rev. Canon ATKINSON ; being a Book of Devotions for the Twenty-four Hours. Printed in Red and Black. Fcp. 8vo, cloth, gilt top, 3-r. 6d. Golden Sayings : Being Selections from the Works of JEREMY TAYLOR, arranged and with an Introduction by JOHN DENNIS. Crown 8vo, buckram, gilt top, 35. 6d. Mr. Zinzan of Bath ; or, Seen in an Old Mirror. A Novel. By MARY BATHURST DEANE. With numerous Illustrations by L. DEANE. Fcap 4to, cloth, 6s. Three Little Maids. By MARY BATHURST DEANE. With 29 Illustrations by EDITH ELLISON. Crown 8vo, cloth extra, $s. Also with Frontispiece only in " Tipcat " Series. " That is the worst of this story there is no next volume " Academy. " It is simply one of the most delightful books of the season. We had made up our minds to say as much before reading half of it, and yet went on to the end. A hard-worked reviewer struggling against the torrent of gift-books can hardly pay a better compliment." Spectator. Miss Margery's Ways. By J. A. LEFROY. With Frontispiece by BANNISTER. Small Crown 8vo, cloth, ^s. 6d. Angela. By ALICE WEBER. 4to, in parchment wrapper designed by T. PYM. 2s. 6d. LONDON: A. D. INNES & Co., 31 & 32 BEDFORD ST., STRAND. THE READER'S SHAKESPEARE. New Edition. Complete in Nine Volumes, large crown 8vo, cloth, gilt top, with portrait, 40s. the Set. Single Volumes, $s. each. This edition has been printed from a fount of new type, at the University Press, Oxford. Vols. I. III. contain the Comedies. Vols. IV. V. contain the Histories. Vols. VI. VIII. contain the Tragedies. Vol. IX. Songs, Sonnets, and Poems. ADVERTISEMENT. THE present Work contains the Complete Writings of the Poet, with the omission of the doubtful plays. After careful collation of the standard editions, Emendations (of doubtful or corrupt passages only) have been admitted with rather more freedom than would be allowable in a purely critical text ; and Notes are entirely excluded. For the plan of this Edition combines features which, it is believed, entitle it to a place apart from and hitherto un- occupied by any other. Its distinctive aim is to ensure, in an unusual degree, the comfort of the ordinary reader. With this view, special pains have been taken to render the typo- graphy peculiarly attractive, and the size of the volumes particularly convenient. It is hoped that, with these qualifications, the READER'S SHAKESPEARE will be found to justify its name. LONDON : A. D. INNES Co., 31 & 32 BEDFORD ST., STRAND. THE DAINTY BOOKS. Dainty 4/0 (5^ x 5), Uniform, Illustrated, Artistically Bound, with Gilt Top, each Volume 2s. 6d. Under this title Messrs. A. D. 1NNES &> CO. are issuing a series of Volumes in a form at once novel and convenient, and specially designed to deserve their title in every respect. The series is intended for children of all ages. Each Volume contains mimerous illustrations. For Grown-up Children. By Mrs. WALFORD. With illustrations by T. PYM. Mum Fidgets. By CONSTANCE MILMAN, Author of " The Doll Dramas. " With illustrations by T. PYM. Master Bartlemy. By FRANCES E. CROMPTON, Author of " Friday's Child." With illustrations by T. PYM. Lily and Water-Lily. By Mrs. COMYNS CARR. With illustra- tions by WINIFRED SMITH. A Mannerless Monkey. By MABEL WOTTON. With illustra- tions by EDITH ELLISON. A Hit and A Miss. By the Hon. E. KNATCHBULL-HUGESSEN. With illustrations by L. LESLIE BROOKE. THE ROSELEAF LIBRARY. A series of books for children, charmingly bound and profusely illustrated, large crown %vo, cloth, $s. bd. each Volume. By STELLA AUSTIN. Paul's Friend. With 16 illustrations by S. B. GATES. By ROMA WHITE. Brownies and Boseleaves. With 19 illustrations by L. LESLIE BROOKE, Cover designed by the Artist. By F. E. CROMPTON, Author of " Master Bartlemy." The Gentle Hermitage. With 27 illustrations by T. PYM. Cover designed by the Artist. By E. N. LEIGH FRY. Shreds and Patches. With 8 full-page illustrations by EDITH ELLISON. By E. PUTNAM and ARLO BATES. Prince Vance. With 32 illustrations by FRANK MYRICK. LONDON: A. D. INNES & Co., 31 & 32 BEDFORD ST., STRAND. THE 'TIP-CAT SERIES. Each Volume uniform, cloth extra, gilt edges, with Frontispiece, $s. 6d. A Select Series of Books for Girls, uniform in style, binding, and price. By the Author of " TIP-CAT " and " Miss TOOSEY'S MISSION." Dear. in. Our Little Ann. Pen. Tip-Cat. By HELEN SHIPTON. Dagmar. By C. M. YONGE. Countess Kate and the Stokesley Secret. By STELLA AUSTIN. Little Princess Angel. By the Author of " MADEMOISELLE MORI." Lilian and Lili. By ALICE WEBER. Leal Souvenir. By C. R. COLERIDGE and M. BRAMSTON. Truth with Honour. By L. T. MEADE. A Ring of Rubies. By MARY BATHURST DEANE. Three Little Maids. LONDON : A. D. INNES & Co., 31 & 32 BEDFORD ST., STRAND. AIDS TO DEVOTION. PRICE HALF-A-CROWN A VOLUME. A. Series of Devotional Works by the Bev. E. B.'.PTTSEY, Canons JELF and ASH WELL, &c., &c. Crown or Fcap. 8vo, well bound in cloth. the Catholic Faith. Notes and Questions. Complied from the Works of the Rev. E. B. PUSEY. With Preface by Canon CARTER. Lent lectures on the Holy Catholic Church. By the Rev. A. R. ASHWELL. Lent Readings from the Fathers. By the Rev. E. B. PUSEY. The Gospel and the Home. Readings for Busy People. By CAROLINE M. HALLETT. Spiritual Combat, with the Path of Paradise. By SCUPOLI. Edited by the Rev. E. B. PUSEY. The Life of Jesus Christ in Glory. By NOUET. Edited by the Rev. E. B. PUSEY. The Sufferings of Jesus. Edited by the Rev. E. B. PUSEY. 2 VOls. * Foundations of the Spiritual Life. By SURIN. Edited by the Rev. E. B. PUSEY. The Meditations of St. Anselm. By ANSELM. Edited by the Rev. E. B. PUSEY. The Consolations of the Christian Seasons. By the Rev. G. E. JELF. Ascension to All Saints. Make up for Lost Time. By the Rev. G. E. JELF. Hear the Church. Lectures on Church Principles. By the Rev. G. E. JELF. Spiritual Readings from Jeremy Drexelius. Transl ated by the Author of " Charles Lowder." Edited by the Rev. W. H. CLEAVER. Sermons on the Litany and the Church. By the Rev. JOHN KEBLE. Companion to the Almanack and Church Calendar. By Mrs. VERNON DELVES BROUGHTON. Advent Readings from the Fathers. By the Rev. E. B. PUSEY. A Guide for Passing Advent Holily. By AVRILLON. Edited by the Rev. E. B. PUSEY. LONDON: A. D. INNES & Co., 31 & 32 BEDFORD ST., STRAND. K THIS BOOK 1HE LAST OVERDUE. 181837 DEC 22 1342 8Jul'49JS ?!