Meanwhile the light diminished sensibly, surrounding objects appearing as if seen by moonlight. The decisive moment was near, and we waited for it with some anxiety; this anxiety took nothing from our powers of observation, it rather stimulated and increased them; and it was, besides, fully justified by the grandeur of the spectacle which nature was preparing for us, and by the consciousness that the fruits of our thorough preparations and of a long voyage would depend on the use now made of a few minutes.
The solar disc was soon reduced to a narrow bright arc, and we redoubled our attention. The slits of the spectroscopes were kept precisely upon the part of the moon's limb where the last light of the sun would be seen, so that they would be directed to the lower regions of the solar atmosphere at the moment of contact of the discs.
The total obscuration occurred instantaneously, and the spectral phenomena also changed immediately in a very remarkable manner. Two spectra, formed of five or six very bright lines—red, yellow, green, blue, and violet—occupied the field in place of the prismatic image of the sun which had just disappeared. These spectra, about one minute (of arc) long, corresponded line for line, and were separated by a dark space in which I could see no lines.
The finder showed that these two spectra were caused by two magnificent protuberances which were now visible on each side of the point of contact. One of them, that on the left, was more than three minutes (or one tenth of the sun's diameter) in height; it looked like the flame of a furnace, rushing violently from the openings of the burning mass within, and driven by a strong wind. The one to the right presented the appearance of a mass of snowy mountains, with its base resting on the moon's limb, and enlightened by a setting sun. These appearances have been carefully drawn by M. Jules Lefaucheur. I will therefore only remark before quitting the subject, which I shall have to treat subsequently under a special aspect, that the preceding observation shows at once:
1st. The gaseous nature of the protuberances, (the lines being bright.)
2d. The general similarity of their chemical composition, (the spectra corresponding line for line.)
3d. Their chemical species, (the red and blue lines of their spectrum being no other than the lines C and F of the solar one, and belonging, as is well known, to hydrogen gas.)
Let us now return to the dark space which separated the spectra of these protuberances. It will be remembered that, at the moment of the total obscuration, the slits were tangent to the solar and lunar discs, and were therefore directed toward the circumsolar regions immediately above the photosphere, in which regions M. Kirchhoff's theory places the atmosphere of vapors, which produces by absorption the dark lines of the solar spectrum. This atmosphere, when shining by its own light, should, according to the same theory, give a reversed solar spectrum, that is to say, one composed entirely of bright lines. This is what we were expecting and trying to verify, and it was to make the proof decisive that I had used so many precautions. But we have just seen that only the protuberances gave positive or bright-line spectra. Now, it is very certain that, if an atmosphere formed of the vapors of all the substances which have been found in the sun really existed above the photosphere, it would have given a spectrum at least as brilliant as that of the protuberances, which were formed of a gas much less dense and less luminous. It must, then, be admitted that, if this atmosphere exists, its height is so small that it has escaped notice.
I must also add that this result did not much surprise me; for my investigations on the solar spectrum had led me to doubt the reality of any considerable atmosphere around the sun, and I am more and more inclined to think that the phenomena of elective absorption, ascribed by the great physicist of Heidelberg to an atmosphere exterior to the sun, are clue to the vapors of the photosphere itself, in which the solid and liquid particles forming the luminous clouds are floating. This view is not merely in harmony with the beautiful theory on the constitution of the photosphere which we owe to M. Faye, but even seems to be a necessary deduction from it.
In fine, the eclipse of the 18th of August appears to me to show that the formation of the solar spectrum cannot be explained by the theory heretofore admitted, and I propose a correction to this theory as above indicated.