49. Here, then, there can be no doubt that the stars are connected, and such pairs are called physical couples, to distinguish them from the optical couples, in which the component stars are really distant from each other, and have no real connexion; their apparent nearness to each other being an appearance caused by their lying in the same straight line, as seen from the Earth.

50. Where the distance of a physical double star is known, we can determine the dimensions of the orbit of one star round the other, as we can determine the Earth's orbit round the Sun. Thus we know that the distance between the two stars of 61 Cygni is 4,275,000,000 miles, and yet the two stars seem as one to the naked eye.

51. The stars are not only of different magnitudes (Art. 22), but the brilliancy of some particular stars changes from time to time. If the variation in the light is, as it is generally, slow, regular, and within certain limits, stars in which this is noticed are called variable stars, or shortly, variables. In some cases, however, the increase and decrease have been sudden, and in others the limits of change have been unknown; and hence we read of new stars, lost stars, and temporary stars, in addition to the more regular variables. There is little doubt, however, that all these phenomena are the same in kind, though different in degree.

52. The variation is, of course, determined by the different magnitudes of the stars at different times, and the amount of variability is measured by the extreme magnitudes. The period of the variability is the time that elapses between two successive greatest bright

nesses.

53. We give a table of a few variable stars, in order that the foregoing may be clearly understood :

54. The fourth star on our list is a very interesting one, for at its period of greatest brightness it sometimes reaches the first magnitude, sometimes the second; but among the acknowledged variables ẞ Persei is perhaps the most interesting, as its period is so short, and, unlike o Ceti(called also Mira, or the Marvellous)-—it is never invisible to the naked eye. The star in question shines as a star of the second magnitude for two days and thirteen hours and a half, and then suddenly loses its light and in three hours and a half falls to the fourth magnitude; its brilliancy then increases again, and in another period of three hours and a half it reattains its greatest bright-all the changes being accomplished in less than three days.

ness

55. Among the new, or temporary stars, those observed in 1572 and last year (1866) are the most noticeable. The first appeared suddenly in the sky and was visible for seventeen months; its light at first was equal to that of the planets at their greatest brilliancy; so bright was it indeed, that it was clearly visible at noonday. Now it is not a little curious that in the years 945 and 1264 something similar was observed in the same region of the sky (in Cassiopea) in which this star appeared. If then

we assume all these phenomena to be due to the fact that we have here a long-period variable star which is very bright at its maximum and fades out of view at its minimum, we may expect a reappearance of the star in the year 1885.

56. We now come to the new star which broke upon our sight last year, in the constellation of Corona Borealis, and which was observed with much minuteness and with powerful methods of research not employed before. This star was recorded some years ago as one of the ninth magnitude. In May, however, it suddenly flashed up, and on the 12th of that month shone as a star of the second magnitude. On the 14th it had descended to the third magnitude,—the decrease of brightness was for some time at the rate of about half a magnitude a day, and towards the end of May it was less rapid. There is good reason to believe that this increased brilliancy was due to the sudden ignition of hydrogen gas in the star's atmosphere. Here we have a fact which must prove of the highest importance, although we are not yet in a position to do much more than speculate upon it.

57. The question of variable stars is one of the most puzzling in the whole domain of astronomy. Mr. Balfour Stewart, from his researches on the Sun-which is doubtless a variable star thinks that ". we are entitled to conclude that, in our own system, the approach of a planet to the Sun is favourable to increased brightness, and especially in that portion of the Sun which is next the planet." In the case of variable stars, the hypothesis which was formerly thought to give the best explanation of the phenomenon is that which assumes rotation on an axis, while it is supposed that the body of the star is not equally luminous on every part of its surface

58. If, instead of this, we suppose such a star to have a large planet revolving round it at a small distance, then,

according to Mr. Stewart's theory, that portion of the star which is near the planet will be more luminous than that which is more remote; and this state of things will revolve round as the planet itself revolves, presenting to a distant spectator an appearance of variation, with a period equal to that of the planet.

59. If we suppose the planet to have a very elliptical orbit, then for a long period of time it will be at a great distance from its primary, while, for a comparatively short period, it will be very near. We should, therefore, expect a long period of darkness, and a comparatively short one of intense light-precisely what we have in temporary

stars.

LESSON IV.-COLOURED STARS. APPARENT SIZE. THE STRUCTURE OF THE STARS. CLUSTERS OF STARS.

60. The stars shine out with variously coloured lights; thus we have scarlet stars, red stars, blue and green stars, and indeed stars so diversified in hue that observers attempt in vain to define them, so completely do they shade into one another. Of large stars of different colours we may give the following table, founded on Mr. Ennis's observations :—

61. In the double and multiple stars, however, we meet with the most striking colours and contrasts; Iota (1) Cancri, and Gamma (y) Andromeda, may be instanced. In Eta (n) Cassiopea we find a large white star with a rich ruddy purple companion. Some stars occur of a red colour, almost as deep as that of blood. What wondrous colouring must be met with in the planets lit up by these glorious suns, especially in those belonging to the compound systems, one sun setting, say in clearest green, another rising in purple or yellow or crimson; at times two suns at once mingling their variously coloured beams! A remarkable group in the Southern Cross produced on Sir John Herschel "the effect of a superb piece of fancy jewellery." It is composed of over 100 stars, seven of which only exceed the tenth magnitude; among these, two are red, two green, three pale green, and one greenish blue.

62. The colours of the stars also change. If we go back to the times of the ancients, we read that Sirius, which is now green, was red; that Capella, which is now pale blue, was also red.

63. In some variable stars the changes of colours observed are very striking. In the new star of 1572, Tycho Brahe observed changes from white to yellow, and then to red; and we may add that generally when the brightness decreases the star becomes redder.

64. The size or diameter of the stars cannot be determined by our most powerful instruments; but we know that, as seen from the Earth, they are, in consequence of their distance, mere points of light, so small as to be beyond all our most delicate measurement. The Moon, which travels very slowly across the sky, sometimes (as we shall see by and by) gets before, or eclipses, or occults, some of them; but they vanish in a moment-which they would not do if they were not as small as we have stated.