the sun, might exercise such an action; and he showed that the solar spectrum contained a whole system of dark and fine lines, comparable to the lines of solar origin, but which were due to the action of our atmosphere. Brewster had already discovered that the solar spectrum was enriched with dark bands at sunrise and sunset, but that in his instrument they wholly disappeared during the day. Both Brewster and Gladstone, his eminent co-laborer, declared in their last memoir on this subject, in 1860, that they could not determine the cause of the phenomenon.
The absorbing action of our atmosphere was still more plainly demonstrated by an experiment on the Lake of Geneva, in which the absorption rays were obtained with the light of a fire passing over Lake Leman, from a distance of fourteen miles. It was also shown in an experiment made at Villette, with a tube filled with vapor at seven atmospheres of pressure and one hundred and twenty feet long, that the vapor of water has a complete absorption spectrum, and that the largest proportion of the absorption phenomena of our atmosphere should be attributed to it.
These observations and experiments doubled the field of investigation opened to spectrum analysis. Not only could the incandescent atmospheres of the sun and the stars now be made to reveal their nature and their composition to us, but our researches might also be extended to objects having a still greater interest for us. We could at once take our own atmosphere for an object, investigating high and inaccessible regions, and making analyses in them which could not be attempted by any other means. Then, going away from the earth, we could interrogate the planetary atmospheres, and seek in them the vapor of water, and with it one of the first conditions of the development of terrestrial life. We could also, comparing the composition of the planetary atmospheres with the astronomical facts which permit us to judge of the geological conditions of the surfaces of the planets, follow in them the atmospheric evolutions which on the earth belong to the domain of the past or of the future. Finally, the same study of the planetary atmospheres, when it shall have become more complete, will show us whether our atmosphere is a type reproduced everywhere, the composition of which appears from that fact indispensable to the existence of living beings, or whether, discovering atmospheres of varied compositions, we shall be led to suppose that life may appear and be developed in media essentially different. The planetary stars are not, however, the only ones that lend themselves to these applications. There are also fixed stars the spectra of which reveal the characteristics of the vapor of water. Now, we know that the atmosphere of a star must be considerably cooled to permit the gases of which water is constituted to combine and generate a vapor. Our sun is still very far from this critical condition. It is also remarkable that the stars presenting these characteristics are generally red or
