LECT. V.] SPECTRUM OF THE CORONA. 269 Watson's sketch (No. 1) of the same eclipse as seen at Carlentini, which represents the main features of the three pictures. We may sum up as follows the evidence which we at present possess concerning the corona: The corona is due in great part to the sun. (1) Because the whole appearance and structure of the corona indicates a solar origin, and cannot be explained by terrestrial glare. (2) Because the brightness at the moon's centre during a total eclipse without cloud is the measure of the greatest possible glare, and the photograph of Mr. Brothers shows how small this is. (3) Because the dark rifts are solar phenomena. There is, however, little doubt that, although the greater portion of the coronal light is due to incandescent gas, giving the bright lines, a certain part is derived from other sources. Thus a faint continuous spectrum would indicate the presence of incandescent cosmical dust; the appearance of the dark Fraunhofer's lines shows that ordinary solar light is reflected; whilst the fainter and outlying part may be due to light reflected from our own atmosphere. It is very singular that the positions of these three lines (1250, 1350, and 1474, on Kirchhoff's scale) coincide, within the probable errors of observation, with three lines observed by Professor Winlock in the spectrum of the aurora borealis. The line 1474 is due to iron. Can we suppose that Dalton's old speculations as to the nature of the aurora have after all some foundation, and that this beautiful phenomenon is due "to the presence of some elastic fluid substance, probably of a ferruginous nature, existing in the higher regions of our atmosphere?" 1 Ångström also observed the spectrum of the aurora in the winter of 1867-8; he saw only one very bright band, the wave-length of which he finds is 5567 tenmillionths of a millimetre, or at 1259 (between D and E) on Kirchhoff's scale, together with traces of two others: but the most interesting discovery which he made on this branch of the subject was, that he succeeded in observing the same bright band in the spectrum of the zodiacal light; and even on a starlight night, when the whole sky seemed almost phosphorescent, Ängström saw traces of these bands in light from all parts of the heavens. In the last few years several splendid exhibitions of aurora have been witnessed: those who saw the display of 25th October, 1870, can never forget the effect. In this and other auroras the same lines were seen; thus Winlock in 1869 saw five bright lines, one of which is identical in position with the coronal line 1474 (Kirchhoff), and Zöllner and others in 1870 observed the same lines. Strangely enough, this band, which links together the apparently unconnected phenomena of the solar corona, the zodiacal light, and the aurora, does not, according to Ångström, appear to correspond to the lines of any known substance. We must, however, remember that if the terrestrial aurora is caused by particles of the carth's atmosphere, and that if it sometimes occurs at an elevation of more than 100 miles above the earth's surface, it follows that the light has been able to pass through 100 miles of cold atmosphere without being altogether absorbed. The radiating power of the atmosphere for these rays must, therefore, be very small, since its absorbent power is also very small. Hence, then, we ought not to be surprised if we cannot reproduce these 1 See end of Appendix A. |