distance in sufficient quantity, and at a luminous temperature, itmay be possible at other times to detect their light with a spec-troscope by carefully inspecting the circumference of the Sun at a moderate distance from the visible boundary of its disc.There will be much brilliant glare of ordinary solar lightarising from the illumination of the Earth ’s atmosphere im-mediately around the direction in which the Sun is seen, whichwill be greatly in our way. But the principle of the con-struction of the spectroscope is such, that, if we increase thenumber of prisms used in it, we may lengthen out the spec-trum of that glare more and more, and make it increasinglyfaint.
At the same time, if we are looking at the light of any massof vapour rising up from the edge of the Sun , the bright linescaused by that vapour will not get fainter as we increase ourbattery of prisms, but will only be further separated from oneanother. It is consequently found that when we direct ourinstrument to the neighbourhood of the Sun ’s apparent edge,we can often detect the bright lines belonging to masses ofvapour rising up from it, or suspended over it. We are thusable to view, not only during a total eclipse, but at any time ,,much of what is going on in regions where it might have beensupposed that the blinding brilliancy of the adjacent photo-spheric light would have for ever utterly baffled our observations.The credit of this discovery, which was made in the year 1868,is due to Messrs. Janssen and Lockyer.
It would, however, be a very laborious process to find outthe shape and boundaries of any such masses of vapour byslowly moving the spectroscope about, and noticing exactlywhen the bright lines corresponding to the vapour in questionmight appear or disappear, as the field of view of the instru-ment might pass across the vapour at different levels insuccession. By so doing we might, no doubt, obtain somesuch result as is indicated in Fig . VIII., in which a series ofsw'eeps in directions parallel to the Sun ’s edge are supposed tobe made, and the general shape and extent of the vapour tobe approximately deduced by the length of the parallel brightlines successively seen. For the sake of contrast, the sur-