SUN 475 during eclipse, precisely as it is illuminated by the sun when there is no eclipse ; but it will readily be understood that this portion of re- flected light is very small in amount.) During the solar eclipse of August, 1869, Profs. Young and Harkness discovered that certainly one bright line exists in the spectrum of the corona, and two other lines were suspected. European astronomers expressed doubt as to the accuracy of this observation ; but it was confirmed du- ring the Mediterranean eclipse of December, 1870, when Young thus summed up his own and other observations : " There is surrounding the sun, beyond any further reasonable doubt, a mass of self-luminous gaseous matter, whose spectrum is characterized by the green line 1,474 Kirchhoff. The precise extent of this it is hardly possible to consider as determined, but it must be many times the thickness of the red hydrogen portion of the sierra, perhaps on an average 8' or 10', with occasional horns of twice that height. It is not at all unlikely that it may even turn out to have no upper limit, but to extend from the sun indefinitely into space." During the same eclipse, Brothers of Manchester and Willard of Philadelphia (the latter acting under the directions of Prof. Win- lock of the Harvard observatory) photographed the corona successfully from two distant sta- tions', Willard being near Jerez in Spain, Broth- ers near Syracuse in Sicily. The views thus obtained agreed so closely (save in circum- stances depending on photographic conditions) as to leave no doubt that the corona is a solar phenomenon. Doubts were still expressed, and it was not until the solar eclipse of De- cember, 1871, that these were finally removed. On that occasion the spectroscopic and pho- tographic results were alike decisive. Jans- sen with the spectroscope not only recognized the bright lines before seen and others less bright, but also a faint solar spectrum, which, since our atmosphere during total eclipse is certainly not illuminated by sunlight, must have been reflected by matter in the solar corona, such as vaporous clouds, meteor flights, or the like. Mr. Davis, a photographer sent out at Lord Lindsay's expense, obtained five excellent photographs of the corona, all agree- ing perfectly inter se, excepting in extent. This proved certainly that the features of the corona do not change as they would if the phenomenon depended on the passage of light rays athwart lunar inequalities, to fall upon scattered matter at a less distance than the moon. Again Col. Tennant obtained six pho- tographs, similarly accordant inter se, and also agreeing perfectly with Mr. Davis's at Doda- betta, a station far removed from Davis's, Bai- cull. Since, also, Dodabetta is near the highest peak of the Neilgherries, about 9,000 ft. above the sea level, while Baicull is close to the sea- shore, it will be manifest that if the features of the corona depended on the illumination of our own atmosphere, the pictures of Tennant's series would have differed altogether from those of Davis's series. Thus, independently of the spectroscopic evidence, the photographs proved that the corona is a solar appendage, at least as far as those features shown in the two series extend. But they extend from the sun in places to a distance exceeding his own diame- ter, and amounting in fact to more than a million miles. There is reason to believe that the true solar corona extends much further, and that in reality the zodiacal light (see ZODI- ACAL LIGHT) forms the outer part of the solar corona; so that if the light of the sun could be for a time obliterated without rendering his appendages invisible, we should see the corona merging gradually into the faint glow of the zodiacal light. Mr. Arthur W. Wright of Yale college has succeeded in showing that this light is not emitted from incandescent gas, but reflected from particles or small bod- ies, and hence derived from the sun. Another important discovery made during total solar eclipses relates to a solar atmosphere under- lying even the sierra. Secchi had observed in 1869 that close to the sun's limb the solar spectrum becomes continuous; this he con- sidered to be due to the existence of a rela- tively very shallow atmosphere, consisting of the vapors which cause the dark lines of the solar spectrum. For if the brightness of the lines of these vapors corresponds very closely to the brightness of the ordinary solar spec- trum for the parts near to the sun's edge, the dark lines of the latter spectrum would be cancelled, and so a continuous spectrum would be produced. For another reason, the present writer had adopted the theory that the atmos- phere producing the absorption lines of the solar spectrum must be shallow, compared at least with the dimensions of the sun's globe ; for he showed that a shallow and not a deep atmosphere is to be inferred from the darken- ing of the solar disk near its edge. The opin- ion thus advanced on theoretical grounds was shown to be correct by the observations of Prof. Young during the total eclipse of De- cember, 1870 ; for, " directing his analyzing spectroscope to the part of the sun's limb which was to disappear last, he found that at the instant when totality commenced the solar spectrum was suddenly replaced by a spectrum consisting of a thousand soft bright lines." In other words, the vapors which by their absorp- tive action produce the dark lines of the ordi- nary solar spectrum were for the moment shi- ning with their own light, and thus produced a spectrum of bright lines. This spectrum continued visible for a few seconds only, show- ing that the complex atmosphere producing it cannot be more than two or three hundred miles in depth. The observation was success- fully renewed during the eclipse of December, 1871, and again during the annular eclipse of June, 1872. How to account for the supply of the prodigious amount of heat constantly radiated from the solar surface has offered a boundless field of hypothesis. One conjecture
