horizon. In fact, these changes of colour must not be regarded as indicating aught but the star's superior brightness. Every very bright star, when close to the horizon, shows these colours, and so much the more distinctly as the star is the brighter. Sirius, which surpasses the brightest stars of the northern hemisphere full four times in lustre, shows these changes of colour so conspicuously that they were regarded as specially characteristic of this star, insomuch that Homer speaks of Sirius (not by name, but as the star of autumn) shining most beautifully when laved of ocean's wave-that is, when close to the horizon. And our own poet, Tennyson, following the older poet, sings how

the fiery Sirius alters hue,

And bickers into red and emerald.

The new star was brighter than Sirius, and was about five degrees lower down, when at its highest above the horizon, than Sirius when he culminates. Five degrees being equal to nearly ten times the apparent diameter of the moon, it will be seen how much more favourable the conditions were in the case of Kepler's star for those coloured scintillations which characterised that orb. Sirius never rises very high above the horizon. In fact, at his highest (near midnight in winter, and, of course, near midday in summer) he is about as high above the horizon as the sun at midday in the first week in February. Kepler's star's greatest height above the horizon was little more than three-fourths of this, or equal to about the sun's elevation at midday on January 13 or 14 in any year.

Like Tycho Brahe's star, Kepler's was brighter even than Jupiter, and only fell short of Venus in splendour. It preserved its lustre for about three weeks, after which time it gradually grew fainter and fainter until some time between October 1605 and February 1606, when it disappeared. The exact day is unknown, as during that interval the constellation of the Serpent-Bearer is above the horizon in the daytime only. But in February 1606, when it again became possible to look for the new star in the night-time, it had vanished. It probably continued to glow with sufficient lustre to have remained visible, but for the veil of light under which the sun concealed it, for about sixteen months altogether. In fact, it seems very closely to have resembled Tycho's star, not only in appearance and in the degree of its greatest brightness, but in the duration of its visibility.

In the year 1670 a new star appeared in the constellation Cygnus, attaining the third magnitude. It remained visible, but not with this lustre, for nearly two years. After it had faded

almost out of view, it flickered up again for a while, but soon after it died out, so as to be entirely invisible. Whether a powerful telescope would still have shown it is uncertain, but it seems extremely probable. It may be, indeed, that this new star in the Swan is the same which has made its appearance within the last few weeks; but on this point the evidence is uncertain.

On April 28, 1848, Mr. Hind (Superintendent of the Nautical Almanac, and discoverer of ten new members of the solar system) noticed a new star of the fifth magnitude in the Serpent-Bearer, but in quite another part of that large constellation than had been occupied by Kepler's star. A few weeks later, it rose to the fourth magnitude. But afterwards its light diminished until it became invisible to ordinary eyesight. It did not vanish utterly, however. It is still visible with telescopic power, shining as a star of the eleventh magnitude, that is, five magnitudes below the faintest star discernible with the unaided eye.

This is the first new star which has been kept in view since its apparent creation. But we are now approaching the time when it was found that as so-called new stars continue in existence long after they have disappeared from view, so also they are not in reality new, but were in existence long before they became visible to the naked eye.

On May 12, 1866, shortly before midnight, Mr. Birmingham, of Tuam, noticed a star of the second magnitude in the Northern Crown, where hitherto no star visible to the naked eye had been known. Dr. Schmidt, of Athens, who had been observing that region of the heavens the same night, was certain that up to 11 P.M., Athens local time, there was no star above the fourth magnitude in the place occupied by the new star. So that, if this negative evidence can be implicitly relied on, the new star must have sprung at least from the fourth, and probably from a much lower magnitude, to the second, in less than three hourseleven o'clock at Athens corresponding to about nine o'clock by Irish railway time. A Mr. Barker, of London, Canada, put forward a claim to having seen the new star as early as May 4-a claim not in the least worth investigating, so far as the credit of first seeing the new star is concerned, but exceedingly important in its bearing on the nature of the outburst affecting the star in Corona. It is unpleasant to have to throw discredit on any definite assertion of facts; unfortunately, however, Mr. Barker, when his claim was challenged, laid before Mr. Stone, of the Greenwich Observatory, records so very definite of observations made. on May 4, 8, 9, and 10, that we have no choice but either to admit these observations, or to infer that he fell under the

delusive effects of a very singular trick of memory. He mentions in his letter to Mr. Stone that he had sent full particulars of his observations on those early dates to Professor Watson, of Ann Arbor University, on May 17; but (again unfortunately) instead of leaving that letter to tell its own story in Professor Watson's hands, he asked Professor Watson to return it to him: so that when Mr. Stone very naturally asked Professor Watson to furnish. a copy of this important letter, Professor Watson had to reply, 'About a month ago, Mr. Barker applied to me for this letter, and I returned it to him, as requested, without preserving a copy. I can, however,' he proceeded, 'state positively that he did not mention any actual observation earlier than May 14. He said he thought he had noticed a strange star in the Crown about two weeks before the date of his first observation-May 14-but not particularly, and that he did not recognise it until the 14th. He did not give any date, and did not even seem positive as to identity. . . . When I returned the letter of May 17, I made an endorsement across the first page, in regard to its genuineness, and attached my signature. I regret that I did not preserve a copy of the letter in question; but if the original is produced, it will appear that my recollection of its contents is correct.' I think no one can blame Mr. Stone, if, on the receipt of this letter, he stated that he had not the slightest hesitation' in regarding Mr. Barker's earlier observations as not entitled to the slightest credit.'1

...

6

It may be fairly taken for granted that the new star leapt very quickly, if not quite suddenly, to its full splendour. Birmingham, as we have seen, was the first to notice it, on May 12. On the

1 Still, a circumstance must be mentioned which tends to show that the star may have been visible a few hours earlier than Dr. Schmidt supposed. Mr. M. Walter, surgeon of the 4th regiment, then stationed in North India, wrote (oddly enough, on May 12, 1867, the first anniversary of Mr. Birmingham's discovery) as follows to Mr. Stone:-'I am certain that this same conflagration was distinctly perceptible here at least six hours earlier. My knowledge of the fact came about in this wise. The night of the 12th of May last year was exceedingly sultry, and about eight o'clock on that evening I got up from the tea-table and rushed into my garden to seek a cooler atmosphere. As my door opens towards the east, the first object that met my view was the Northern Crown. My attention was at once arrested by the sight of a strange star outside the crown' (that is, outside the circlet of stars forming the diadem, not outside the constellation itself). The new star was then certainly quite as brightI rather thought more so-as its neighbour Alphecca,' the chief gem of the Crown. 'I was so much struck with its appearance, that I exclaimed to those indoors, "Why, here is a new comet!' He made a diagram of the constellation, showing the place of the new star correctly. Unfortunately, Mr. Walter does not state why he is so confident, a year after the event, that it was on the 12th of May, and not on the 13th, that he noticed the new star. If he fixed the date only by the star's appearance as a secondmagnitude star, his letter proves nothing; for we know that on the 13th it was still shining as brightly as Alphecca, though on the 14th it was perceptibly fainter.

evening of May 13, Schmidt of Athens discovered it independently, and a few hours later it was noticed by a French engineer named Courbebaisse. Afterwards, Baxendell, of Manchester, and others, independently saw the star. Schmidt, examining Argelander's charts of 324,000 stars (charts which I have had the pleasure of mapping in a single sheet) found that the star was not a new one, but had been set down by Argelander as between the ninth and tenth magnitudes. Referring to Argelander's list, we find that the star had been twice observed-viz., on May 18, 1855, and on March 31, 1856.

Birmingham wrote at once to Mr. Huggins, who, in conjunction with the late Dr. Miller, had been for some time engaged in observing stars and other celestial objects with the spectroscope. These two observers at once directed their telescope armed with spectroscopic adjuncts-the telespectroscope is the pleasing name of the compound instrument-to the newThe result was rather startling. It may be well, however, before describing it, to indicate in a few words the meaning of various kinds of spectroscopic evidence.

comer.

The light of the sun, sifted out by the spectroscope, shows all the colours but not all the tints of the rainbow. It is spread out into a large rainbow-tinted streak, but at various places (a few thousand) along the streak there are missing tints; so that in fact the streak is crossed by a multitude of dark lines. We know that these lines are due to the absorptive action of vapours existing in the atmosphere of the sun, and from the position of the lines we can tell what the vapours are. Thus, hydrogen by its absorptive action produces four of the bright lines. The vapour of iron is there, the vapour of sodium, magnesium, and so on. Again, we know that these same vapours, which, by their absorptive action, cut off rays of certain tints, emit light of just those tints. In fact, if the glowing mass of the sun could be suddenly extinguished, leaving his atmosphere in its present intensely heated condition, the light of the faint sun which would thus be left us would give (under spectroscopic scrutiny) those very rays which now seem wanting. There would be a spectrum of multitudinous bright lines, instead of a rainbow-tinted spectrum crossed by multitudinous dark lines. It is, indeed, only by contrast that the dark lines appear dark, just as it is only by contrast that the solar spots seem dark. Not only the penumbra but the umbra of a sunspot, not only the umbra but the nucleus, not only the nucleus but the deeper black which seems to lie at the core of the nucleus, shine really with a lustre far exceeding that of the electric light, though by contrast with the rest of the sun's surface the penumbra looks dark, the umbra darker

still, the nucleus deep black, and the core of the nucleus jet black. So the dark lines across the solar spectrum mark where certain rays are relatively faint, though in reality intensely lustrous. Conceive another change than that just imagined. Conceive the sun's globe to remain as at present, but the atmosphere to be excited to many times its present degree of light and splendour: then would all these dark lines become bright, and the rainbow-tinted background would be dull or even quite dark by contrast. This is not a mere fancy. At times, local disturbances take place in the sun which produce just such a change in certain constituents of the sun's atmosphere, causing the hydrogen, for example, to glow with so intense a heat that, instead of its lines appearing dark, they stand out as bright lines. Occasionally, too, the magnesium in the solar atmosphere (over certain limited regions only, be it remembered) has been known to behave in this manner. It was so during the intensely. hot summer of 1872, insomuch that the Italian observer Tacchini, who noticed the phenomenon, attributed to such local overheating of the sun's magnesium vapour the remarkable heat from which we then for a time suffered.

Now, the stars are suns, and the spectrum of a star is simply a miniature of the solar spectrum. Of course, there are characteristic differences. One star has more hydrogen, at least more hydrogen at work absorbing its rays, and thus has the hydrogen lines more strongly marked than they are in the solar spectrum. Another star shows the lines of various metals more conspicuously, showing that the glowing vapours of such elements, iron, copper, mercury, tin, and so forth, either hang more densely in the star's atmosphere` than in our sun's, or, being cooler, absorb their special tints more effectively. But speaking generally, a stellar spectrum is like the solar spectrum. There is the rainbow-tinted streak, which implies that the source of light is glowing solid, liquid, or highly compressed vaporous matter, and athwart the streak there are the multitudinous dark lines which imply that around the glowing heart of the star there are envelopes of relatively cool vapours.

We can understand, then, the meaning of the evidence obtained from the new star in the Northern Crown.

In the first place, the new star showed the rainbow-tinted streak crossed by dark lines, which indicated its sun-like nature. But, standing out on that rainbow-tinted streak as on a dark background, were four exceedingly bright lines—lines so bright, though fine, that clearly most of the star's light came from the glowing vapours to which these lines belonged. Three of the lines belonged to hydrogen, the fourth was not identified with any

known line.