these three by the new method, when an instrument of adequate dispersive power was employed. The accompanying picture (fig. 72), for example, exhibits the spectrum of a prominence, and of the adjacent portion of the Sun's limb, as shown by Mr. Lockyer's spectroscope. It will be seen that the double line of sodium as well as three lines of the magnesium spectrum are shown. It is only necessary to suppose that M. Rayet saw the two close double lines as single ones in order to account exactly for the nine lines seen by him.*

The following is Fr. Secchi's description of the principal lines seen in the prominence-spectrum under ordinary conditions:-

6

The line c of hydrogen,' he says, 'is the most easily seen of all. It sometimes reaches the enormous height of three minutes, indicating the presence of such colossal prominences as are seen during eclipses. On the limb of the Sun generally the height of this line is very irregular, and on the average attains to from ten to fifteen seconds. This line also extends in a well-marked manner within the limb, overlapping the disc by ten seconds and more. Yet further on the disc there is a region where the line cannot be seen, being neither bright nor dark, but of the same tint exactly as the neighbouring part of the spectrum, which is thus in this part of its length continuous and uniformly bright. With a slit placed parallel to the limb the space throughout which the line disappears is often considerably extended. Outside the disc the line is much brighter near its base than at the summit, and the line is dilated at the base, and seems to terminate in a point where its light fades off, until, as I have already said, it becomes of the same brightness as the neighbouring part of the spectrum before becoming a dark line. Outside the Sun the line is bounded by two dark lines, which appear at first sight to be the effect of contrast, but may probably have another and a real cause. The bright line is often formed of knots and separate pieces, which are evidently so many fragments of different prominences, placed one beyond the other and unequally bright. If the aperture of the slit be enlarged as much as is possible without rendering the light unbearable, the bright line of the rose-coloured fringe is seen to be marked by irregularities which are due to the roughness of texture (scabrosità) of the prominences themselves.'

Mr. Lockyer was presently able to confirm the views of Grant, Secchi, and others, who had, as

After mentioning that he has been able to apply Janssen's method of determining the shape of the prominences by taking line-sections, Secchi remarks that he has often tried to determine whether there is any fixed law as regards the direction in which the prominences are bent, but has hitherto not succeeded in tracing any.

The line in the yellow, near the sodium lines, is about twice and a half as far from the nearest of these as these are from each other. This line is sensibly the prolongation of a bright line in the solar spectrum. In height and brightness it corresponds closely to the c-line, but I have noticed that it will not bear high magnifying power so well. While the F and c lines remain brilliant under such powers, this line becomes weaker, so that only the practised eye can detect it. The line ends in a point, and it often extends itself brilliantly upon the disc of the Sun. Outside the disc it is not bordered by dark lines; on the contrary, with high powers it becomes diffused, and does not stand out sharply like the c-line.

The F-line is in general not so high as the c-line, and grows faint at the extremity, where it takes the form of a lance. Outside the disc it is accompanied by a narrow dark zone on the more refrangible side. Sometimes I have seen it prolonged beyond the edge of the solar disc, as in the case of the c-line; at other times a very fine black thread shows itself on the more refrangible side. This seems to show that the F-line is not due to hydrogen alone.' (This conclusion, however, has been negatived.)

The third line of hydrogen near a I have seen as a bright line, but it is necessary to reduce the dispersive power of the prismatic battery in order to obtain sufficient light.'

Secchi is of opinion that where there are faculæ on or close to the limb, there prominences exist; but he is unable to say with equal confidence that wherever prominences exist there are also faculæ.

He was surprised to find on one occasion, when several lines besides the above-mentioned were visible, that of the three lines forming the group of magnesium one and one only was visible,-a second line holding a position midway between the other two lines of the same group. He remarks that Rayet had observed during the eclipse of August 1868 only two lines of this group, and that these two doubtless corresponded with those seen by himself, of which, as we have seen, one does not accord with either of the two remaining magnesium lines. He was so surprised at this peculiarity that he searched dili

U

already mentioned, enunciated the theory that the Sun is wholly surrounded by a layer or envelope of the coloured matter whence as it would seem the prominences spring.

To whatever part of the Sun's edge he directed his spectroscope the bright lines belonging to the prominence-spectrum could be seen, though reaching but to a short distance from the edge of the Sun, save where there is a prominence. Unaware that this gently for two hours to detect traces of the other lines of magnesium, but could find none whatever. He considers the objections which have been urged against this observation as 'inconclusive-to use no more severe expression-since it is unreasonable to suggest that he would have been careless when observing so remarkable a phenomenon.'

* In Mr. Lockyer's detailed account of his work (Report of the Astronomical Society, February 1869) he states that at the time the spectroscope revealed to him that the prominence-spectrum was never absent, and that, in fact, the prominence-matter formed a continuous envelope round the Sun, he was not aware that such an envelope had been suggested by previous observers.' The account goes on to speak of the physicists who had propounded this theory-Swann, Grant, Von Littrow, Leverrier, and Secchi-describing their views as ideas,' 'suggestions,' and 'surmises,' and adding that 'the experimental proof of the truth of these surmises is due to Mr. Lockyer.' It is bare justice to remark, however, that with some at least of the five physicists abovenamed the theory was much more than a surmise or conjecture. Thus Professor Grant writes, in 1858, 'The zone of a deep red colour observed at Toulon toward the part of the Moon's limb where the Sun was about to emerge, clearly indicates the accumulation of nebulous matter in the lower regions of the solar atmosphere, as well as the condensation of the circumambient fluid of which the latter is composed towards the surface of the Sun, arising from the pressure of the superincumbent strata.' Leverrier wrote in 1860, The existence of a bed of rose-coloured matter, partially transparent, covering the whole surface of the Sun, is a fact established by the observations made during the time of totality in the eclipse of this year.' Secchi was equally convinced of the existence of this rose-coloured region. The observation of eclipses,' he wrote, 'furnishes indisputable evidence that the Sun is really surrounded by a layer of this red matter, of which we commonly see no more than the most elevated points.' If these be surmises, how shall men of science express conviction?

layer had already been discovered, Mr. Lockyer devised a name for it, entitling it the chromosphere.

Whether this envelope is to be regarded as a true solar atmosphere, in the sense in which that term is usually understood, may be seriously questioned. As seen in the telescope it presents a well-defined and very uneven limit, giving the idea rather of a region. of flames or clouds than of an envelope of the nature of an atmosphere. That the substance producing the coloured light of the chromosphere is gaseous admits indeed of no question; but so also the prominences are gaseous. Yet we do not regard these as of the nature of an atmosphere. Now, if the surface of the Sun be covered at all times with small prominences, bearing somewhat the same relation to the gigantic horns' and boomerangs' seen during eclipses that the bushes covering certain forest regions bear to the trees, then there can be no doubt that the chromosphere could not rightly be regarded as an atmosphere. We have

Leverrier and Secchi believed in the existence of this envelope because they had seen it; and Secchi had the further evidence (though no further evidence could be needed) derived from his photographs. Under these circumstances no subsequent observer can claim to have been the first to give experimental proof of the existence of the chromosphere.' Or if,--then Galileo might have claimed to have been the first to give experimental proof of the existence of the Moon or of Jupiter, since undoubtedly he first studied these bodies telescopically. That Mr. Lockyer would have discovered the chromosphere had it not been discovered before (unknown to him), may be well believed, precisely as we may be assured that had not the eclipse observers in 1868 discovered the nature of the prominence-spectrum, Mr. Lockyer would have done That he did not do either of these things is no discredit to his abilities, since no man that ever lived has succeeded in discovering the discovered; and it may be surmised that no man ever will.

80.

no evidence that it even extends downwards as far as the visible limits of the photosphere, since undoubtedly if a distance of forty or fifty, nay, even of two or three hundred miles, separated the lower limit of the chromosphere from the photosphere, no telescopes we possess could suffice (when supplied with suitable spectroscopic appliances) to reveal any trace of this space. A width of two hundred miles at the Sun's distance subtends an arc of less than half a second; and telescopists who know the difficulty of separating a double star whose components lie so close as this will readily understand that a corresponding arc upon the Sun would be altogether unrecognisable.

I am very far, therefore, from accepting with confidence the view that the chromosphere is the true solar atmosphere. Rather I believe it consists of mattergaseous, no doubt, but-suspended in the true solar atmosphere. I think that not only is this view confirmed by the appearance of the chromosphere as depicted in Plate VI., further on, but that the general resemblance as regards structure and appearance between the chromosphere and the prominences suffice to render it at least highly probable that the former can no more be described as the solar atmosphere than the latter. But even more striking is the evidence deduced from Plate V. For here we see in the first figure a prominence which has been obviously erupted. We see it as it springs upward, expanding with the diminishing atmospheric pressure, and we see in the second figure how the erupted matter has slowly begun