In the above estimate, I have supposed the measurement to be made from the sun's visible surface. But it is very unlikely that that surface is the true lower limit of the atmosphere. It seems far more probable that the surface we see is merely a layer of clouds (as Sir William Herschel suggested so long ago) in the solar atmosphere, and that the actual depth of the atmosphere is more truly indicated by the appearances seen when large sun-spots are examined. That these spots are cavities has been abundantly established. That they are openings through layers of solar clouds has not been indeed demonstrated, yet it is difficult to conceive how they can otherwise be interpreted. As to the way in which the spots are formed, theorists are at issue, some urging that there is an uprush from depths beneath the solar surface; others, that there is a downrush of matter from without. But neither of these views is in any way incompatible with Herschel's theory, that the spots are openings in solar cloud-layers.

We might thus be led to compare the solar atmosphere with our own, though it will of course be obvious that there are many marked points of difference. But in our own atmosphere we have at least two distinct cloud-levels, the region, namely, where the cumulus or wool-pack clouds are formed, and that where the cirrus or feathery clouds, make their appearance. There is air above the cirrus clouds, air between the cirrus and cumulus layers, and air between the cumulus clouds and the earth. And precisely in the same way we may conceive that there exists at all times a solar atmospheric region beneath as well as above the cloud-layer which forms the sun's visible surface, and beneath and between the other cloud-layers revealed by telescopic observations.

But passing from the very difficult question suggested by the consideration of regions below the sun's visible surface, let us discuss briefly the bearing of Professor Young's discovery upon our views respecting those outer regions-the coloured prominences and sierra, the corona itself, and, in fine, all the portions of space which lie above the true atmosphere.

In the first place, it seems to me that the observations made during the late eclipse dispose finally of the theory that the coloured sierra is an atmospheric envelope, properly so-called. I had long since been led to question whether the sierra could be so regarded. Let me remind the reader that the sierra is nothing more nor less than the region which Lockyer rediscovered in 1868. It had, in fact, been recognised by telescopists since 1806, the name sierra having been given to it by the observers of the eclipse of 1842. It is a red region, having (as its name implies) a serrated upper surface, as seen in the telescope, and seemingly extending all round the sun's disc. The red prominences appear to spring from its upper surface. Strangely enough, when Lockyer made his ingenious observations of the coloured prominences, he had not heard of this discovery, or had forgotten it. Accordingly, finding traces of prominence-matter all round the sun, he concluded that there was a continuous envelope of hydrogen (mixed with some other gases) surrounding the whole of the sun's globe. It was probably through being misled by this supposition that he gave to the sierra a new name - entitling it the chromosphere - announcing at the same time that its upper surface was smooth in outline. Respighi, the eminent Italian spectroscopist-also working, it would seem, in ignorance or forgetfulness of the prior recognition of the layer-announced presently that the upper surface of the so-called chromosphere 1 was altogether irregular-more irregular, in fact, than the surface of a tempest-tossed sea. On re-examining the sierra, Mr. Lockyer found this to be the case. But perhaps the most striking evidence as to the real aspect of the sierra was afforded during the eclipse of last December, when Fr. Secchi, towards the close of totality, saw around the western half of the moon's disc a complete semicircle of sierra, and noted that this beautiful coloured crescent was formed of multitudes of minute

1 It affords strange evidence of the caution with which new names should be suggested, that this name, embodying, as we see, an erroneous theory, and also perpetuating the remembrance of a mistaken claim, is scarcely yet beginning to fall into disuse. Perhaps its Greek origin and its length may have something to do with this; for although astronomy -at least descriptive astronomy-has hitherto not been disfigured by the hideous nomenclature which botanists and geologists seem to rejoice in, yet there is always a large class of science students who delight in sesquipedal names, as giving an air of profundity to their discourse. It may even be dangerous to hint that the true form of the compound for a colour-sphere is not chromosphere, but chromato-sphere, since the extra syllable will multiply tenfold the favour with which the compound is accepted. When will the tyro learn that the true lover of science 'Projicit ampullas et sesquipedalia verba'?

prominences. This agrees very satisfactorily with my own anticipatory description of the probable nature of the sierra, when I suggested that the sun's surface is probably '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.'

But the larger prominences have been shown by Zollner and Respighi to be phenomena of eruption. They are masses of glowing gas, which have been flung from great depths beneath the visible surface of the sun. May we not conclude that the smaller prominences which constitute the sierra are of like nature ? that they also have been flung from beneath the sun's visible surface ? As respects the larger prominences we can have no manner of doubt, because they have been seen to be flung out in eruptive sort. And this refers to all orders of prominences, except only those very numerous and relatively very small prominences which crowd together so as to form the seemingly continuous coloured sierra. These cannot be watched as the others have been. But it seems highly probable that those among them which are not the remains of loftier prominences, are, like their larger fellows, phenomena of eruption.