Young paid particular attention to the spectroscopic observation of the corona. He also had been misled by erroneous accounts respecting the Indian observations, and so expected to see a faint solar spectrum. He found, on the contrary, that the light of the corona gave a spectrum of bright lines. He saw three such lines, and he considers it certain, from their close agreement with those shown in Professor Winlock's picture of the aurora-spectrum, that the corona is simply an electric discharge, no doubt varying with great rapidity, as we see in the case of the aurora; in fact, that the corona is a permanent solar aurora.' Now, although these accounts seem at first sight discordant, it appears to me that they can be brought into agreement, not only with each other, but with Lieutenant-Colonel Tennant's, by a consideration of the circumstances under which they were severally made. Tennant, seeing only a continuous spectrum, opened the slit somewhat widely: 'the jaws of the slit were wide apart,' he says; too wide, I imagine, to show the bright lines. For, from what is shown at p. 144, it will be seen, that the brightness of the coronal bands 6 method of coincidences has ever yet been attempted. I had written further to the effect that, even when the auroral light is only ruddy to the eye, no red lines are seen, so that we may conclude that the excess of red is due to a peculiarity in the light of mixed refrangibility forming the continuous spectrum,' when I learned that, on the evening of October 24, Mr. J. R. Capron, with one of Browning's small direct-vision spectroscopes (adapted to star observation), had succeeded in observing a line in the red, 'very much like the lithium line, but rather more dusky. It was only well seen in the rosy patches of the aurora, but could be faintly traced wherever the rose-tint at all extended.' or lines could not be increased in this way, though their breadth, and so the total amount of light from them, would be increased in precisely the same proportion as the opening of the slit. But the brightness of the continuous background would be increased in this same proportion. Hence the bright lines which Tennant could not see, on account of their fineness, were changed by opening the slit into broad bands of no greater brightness, and rendered invisible by reason of the increased brightness of the background. An intermediate amount of opening would in all probability have shown the lines. Now we see that Professor Harkness failed even to see a continuous spectrum when he used a narrow slit; and the fineness of the lines (not nearly so brilliant as the prominence-lines) caused them to escape his notice precisely as had happened with Tennant. But when he opened the slit a little,' he saw the continuous spectrum and one bright line. Had he opened it somewhat more, he would not have seen that bright line, but would have failed as Tennant had, and for the same reason. Had he opened it a little less, he would probably have seen the continuous spectrum and the three bright lines, as Professor Pickering did. With a somewhat smaller opening the continuous spectrum would disappear through excessive faintness; but the three bright lines seen by Professor Young would be even more distinctly visible. We see in fact that Professor Young, who succeeded readily in seeing three bright lines, failed to recognise the continuous spectrum. It may be said that this is hypothetical; and so in a sense it is. What an observer would have seen under certain circumstances different from those which actually occurred must necessarily be hypothetical. But as to the matter of fact on which this hypothetical interpretation of the different results is founded, there can be very little question. Professor Young and Pickering saw three bright lines, Professor Harkness saw one such line, and Tennant, Pickering, and Harkness saw a continuous spectrum; while the conditions under which these different results were obtained are known. My interpretation accounts simply and naturally for all the observed spectra. I make the whole question one of slit-opening. With sufficient dispersive power we get, as might be expected-With a very narrow slit, three bright lines (so fine as to be only recognisable on a close scrutiny, such as that given by Professor Young), and the continuous background too faint for recognition. With a slit not quite so narrow, we get a faint continuous spectrum and three lines, still so fine as to require very careful scrutiny for recognition. With a somewhat wider slit, we get a brighter continuous background on which the brightest of the three lines. alone is visible. And lastly, with the jaws of the slit wide apart, we have a yet brighter continuous spectrum, and no visible bright bands. All this is precisely in accordance with what the theory of the spectroscope *With insufficient dispersive power we have a continuous spectrum without bright lines, whether the slit be widely opened or not. requires, and it accords perfectly well with all the observed facts, save one, the failure, on the parts of Harkness and Tennant, to detect the faint bright lines of the corona, when these must have been very fine, owing to the narrowness of the slit. But this will surprise no one familiar with the very varying powers of observers, as respects the recognition of faint objects, or small objects, or objects which are both faint and small. It is, however, always possible, or rather it is highly probable, that different parts of the corona may give different spectra. The ascertained facts are thesethat some parts of the corona do undoubtedly give a spectrum consisting in part of three bright lines; that these lines agree in position with bright lines belonging to the spectrum of the terrestrial aurora; and that, so far as observation has yet gone, the spectrum of the corona contains no dark lines. The direct observations of the corona as seen during the American eclipse were numerous and important. Mr. W. S. Gilman, jun., from whose coloured drawing of the eclipsed Sun Plate VII. is taken, writes thus respecting the appearance of the corona: The general outline of the corona was a tra *This would accord well with what is observed of the spectrum of the aurora borealis. In the communication already referred to (note, p. 338), in which Mr. Birmingham describes his observation of three bright lines in the spectrum of the aurora, he says of the 'intense red of broad areas of light,' that there was here no line whatever to be detected.' And again he adds, the white light seen in some parts of the sky gave only the one principal line in the green.' We have seen also that the red line of the auroral spectrum is not commonly visible. |