So much attention was directed to the solar corona during the discussions which preceded and followed the late eclipse, that a discovery of extreme importance-but not at all associated with the corona-has received far less attention than it deserves. The discovery I refer to is, in fact, more important in its bearing on problems of solar physics than any which has been made since Kirchhoff first told us how to interpret the solar spectrum. It is also intimately connected with the labours of that eminent physicist. I propose briefly to describe the nature of the discovery, and then to discuss some of the results to which it seems to point.

Astronomers have long seen reason to believe that the sun has an atmosphere. And by the word atmosphere I mean something more than mere vaporous or gaseous masses, such as the prominences have been shown to be. A solar envelope, complete and continuous as our own atmosphere, seems undoubtedly suggested by the appearance which the sun's image presents when thrown on a suitably prepared screen in a darkened room; for then the disc is seen to be shaded off continuously towards the edge, where its brilliancy is scarcely half as great as at the centre. The phenomenon is so readily seen, and so unmistakeable, that it is with a sense of wonder one hears that Arago called it in question. To use the words of Sir John Herschel, 'the fact is so palpable that it is a matter of some astonishment that it could ever fail to strike the most superficial observer.' And, again, not only the light but the heat of the outer portions of the sun's image has been estimated. In this case we do not depend upon the perhaps fallible evidence of the eye, but on that of heat-measuring instruments. Fr. Secchi, measuring the heat of different parts of the solar image, has found that of the part near the centre nearly double that from the borders. Lastly, photography gives unmistakable evidence on the subject.

Now, when Kirchhoff discovered the meaning of the solar spectrum, it seemed clear to him that he had determined the nature and constitution of the solar atmosphere. Let us consider the nature of Kirchhoff's discovery.

He found that the dark lines across the rainbow-tinted streak forming the background (as it were) of the solar spectrum, are due to the action of absorbing vapours. The vapours necessarily lie outside the source of that part of the sun's light which produces the rainbow-tinted streak. If those vapours could be removed for a while, we should see a simple rainbow-riband of light. Or if the vapours could be so heated as to be no less hot than the matter beneath them which produces the rainbow spectrum, they would no longer cause any dark lines to appear; but being cooler, and so giving out less light than they intercept, they cut out the dark spaces corresponding to their special absorptive powers. To use Mr. Lockyer's striking, though perhaps not strictly poetical, description of their action, these vapours 'gobble up the light on its way to the observer, so that it comes out with a balance on the wrong side of the account.' Each vapour produces its own special set of lines, occupying precisely those parts of the spectrum which the vapour's light would illuminate if the vapour shone alone. For these vapours, notwithstanding their action in intercepting or absorbing portions of the sunlight, are themselves in reality glowing with a light so intense that the human eye could not bear to rest upon it. If we could examine the vapours we supposed just now removed from the sun, we should obtain the very lines of light which are wanting in the spectrum of the sun.

When Kirchhoff had recognised in this way the presence of absorptive vapours around the real light-globe of the sun, he judged that they form the solar atmosphere. Because, although his mode of observation was not such as to assure him that these vapours completely envelope the sun, yet the telescopic aspect of the sun, and especially that darkening near the edge to which I have just referred, seemed to leave room for no other conclusion. But at this stage of the inquiry Kirchhoff fell into a mistake. He judged that the solar corona was the atmosphere which produced the solar dark lines, as well as the darkening of the sun's disc near the edge. The mistake is one which, as it seems to me, he would have avoided had he taken into account the enormous pressure at which an atmosphere so extensive as the corona would necessarily exist under the influence of the sun's mighty attractive energies. It may easily he shown that if the outer parts of the corona were as rare as the contents of our so-called vacuum-tubes, or even a thousand times rarer, yet according to the laws which regulate atmospheric pressure, the density even at vast heights above the sun's surface would attain to many hundred times that of our heaviest gases. The pressure would, indeed, be so great that we can see no way of escaping the conclusion that, despite the enormous heat, the gases composing the imagined atmosphere would be liquefied or even solidified.

When the observers of the Indian eclipse of 1868 found that the coloured prominences are masses of glowing hydrogen, with other gases intermixed, and when the prominence-spectrum was found to show the hydrogen lines as these appear when hydrogen exists at very moderate pressures, Kirchhoff's view had to be abandoned as altogether untenable. Wherever the vapours exist which produce the solar dark lines, they are undoubtedly not to be looked for in the corona.

But there the lines are. The absorptive action is exerted somewhere. The question is-Where are the absorptive vapours?

At this stage of the inquiry, a very strange view was expressed by Mr. Lockyer - a view which appears to have been founded on a slight misapprehension of the principles of spectrum analysis. He put forward the theory that the absorptive action takes place below the level of the sun's surface as we see it.