If during any total eclipse of the sun, the moon just concealed the whole of the sun's disc (as may well happen), and if our satellite were only complaisant enough to stay still for a few minutes in such a position, so that one of these exact total eclipses could be studied as readily as one of greater extent (which never can happen), then the shallow atmosphere I have been speaking of could be recognised. The difficulty above considered would no longer exist. For the ring of light which actually hides the shallow atmosphere when (he sun is not eclipsed, is an extension of the bright rim of the disc outwards : if the disc is completely hidden, there is no bright rim to be extended, and anything existing close by the sun's globe can be recognised.

But then, unfortunately, no total eclipse can present these desirable features. If a total eclipse is to be worth seeing at all, the moon's disc as seen at the time must be appreciably larger than the sun's. When totality begins the outlines of the two discs just touch at a single point, and when totality ends the two discs just touch at another point; but during all the rest of the totality the two outlines do not touch at all, that of the moon surrounding without touching that of the sun. The outlines of the two discs do twice touch, however, in each case for one moment and at one point. What Professor Young determined to do, therefore, was to bring under special examination that one point where the outlines touch at the exact moment when totality begins. In other words, he directed his special attention to the point where the last trace of the sun's disc was about to disappear. It is perhaps scarcely necessary to say that he did not trust to the powers of his telescope, but that he employed a powerful spectroscope. And further, he did not depend on his own observations alone, but had adjusted a spectroscope for the use of Mr. Pye, an English gentleman residing in the part of Spain where the eclipse-observing parties were stationed, so that that gentleman also might make the required observations.

In his account, Professor Young does not mention what he expected to see. It is probable that he had in his thoughts the observations of Fr. Secchi, and hoped to obtain evidence respecting that shallow atmospheric envelope which Secchi believed in and Lockyer rejected; though it is quite possible he merely desired to ascertain whether the constitution of the lower part of the sierra differed in any marked respect from that of the upper portion. As the moment approached when the last fine sickle of sunlight was to be obscured, the solar spectrum which was visible in the spectroscopic field of view grew rapidly fainter. The region actually examined by Professor Young was in reality a narrow, almost linear space, touching the edge of the sun's disc; so that before totality had commenced he had the light from our own illuminated atmosphere, and not direct sunlight, to deal with. Thus he had just such a solar spectrum as is seen when a spectroscope is directed to the sky in the daytime. But as the moment of totality drew near, the illumination of the atmosphere, and with it the brightness of the rainbow-tinted streak, rapidly diminished. At last the solar spectrum vanished; and then-What was it replaced by? What was found to be the spectrum of the solar atmosphere close by the sun's surface ? In place of the rainbow-tinted riband crossed by thousands and thousands of dark lines, there appeared a new and most beautiful spectruma riband of rainbow-tinted lines, thousands in number and of all degrees of thickness, - hundreds of red lines, and then, in order, hundreds of orange lines, hundreds of yellow, green, indigo, and violet lines, like coloured cross-threads on a black riband, only infinitely more beautiful. A charming spectacle, truly, but so short-lived that no man can ever hope, though he lived to four-score years and ten, to let his eyes rest in all his life for more than ten or twelve seconds on the beautiful array of coloured lines which two men only have as yet beheld. We may increase the dimensions and power of our telescopes until the existence of these lines can be recognised without the aid of eclipse-darkness, but the lines can never be seen, save during eclipse, as Young and his colleague saw them last December. And these observers tell us that in a second or two the lines vanished, the advancing moon hiding the shallow solar atmosphere. If it should ever be given to any man to see six total eclipses (which has never yet happened to any), and to successfully apply in each instance the method employed by Professor Young, then in all, during his life, that man would have seen the beautiful line-spectrum to perfection for some ten or twelve seconds; but not otherwise can even so long a total period of observation be secured. No single observer, then, can hope to learn much about the thousands of lines which have still to be mapped during eclipse opportunities.

But now let us consider the import of the observation. What are these myriads of coloured lines? Every dark line of the solar spectrum, says Professor Young, seemed to have its representative in this bright-line spectrum. Many of the groups of lines which had flashed so quickly into view and endured but so brief a period, were familiar to him; in other words, his study of the solar spectrum had made him conversant with the corresponding groups of dark lines. It follows, then, beyond all possibility of question, that the source of light was a highly complex atmosphere, formed of those very vapours which, by their absorptive power, produce the dark lines-formed, that is, of the vapours of iron and of copper, of zinc, sodium, magnesium, and of all those elements whose presence in the sun's substance had been inferred from the study of the solar spectrum.

Here, then, at length we have the true solar atmosphere, an atmosphere of a highly complex nature, and doubtless exceedingly dense near the visible surface of the sun, because subject to a pressure so enormous. The upper limit of this atmosphere cannot lie very far above the sun's surface, at least not very far compared with the sun's dimensions. Supposing the actual time during which the line-spectrum was visible to have been two seconds, then it is easy to tell how deep the atmosphere is. For in two seconds the moon must have traversed a space corresponding to about three hundred miles at the sun's distance. An atmosphere three hundred miles deep is, therefore, indicated by Professor Young's observations. It need hardly be said, however, that in the excitement of eclipse observation, the estimate of minute intervals of time can scarcely be relied upon, unless checked by instrumental arrangements, which was not the case in the present instance. We may fairly conclude that the depth of the solar atmosphere lies between some such limits as a hundred miles and five hundred miles.