STTN. 697 SUN. directly from such vaporized substances them- selves, if rendered incandescent. Thus the ordi- narj' solar spectrum is crossed by certain well- known dark lines corresponding to the absorption of certain parts of the photosplu'ric light by the absorptive etl'ect of the reversing layer. Now if it were possible in some way to get rid of the light of the photosphere and examine directly the light of the reversing layer, we ought to see the spectrum of onlinarj- incandescent gases. This observation can actually be made at the time of a total solar eclipse. If we watch the solar spectrum very carefully just before the beginning of totality, when the moon's advancing edge is on the point of obscuring the sun com- pletely, there will come a moment when nothing remains in sight but the outer or reversing layer. At that instant the dark lines in the spectrum are suddenly reversed, becoming bright like the lines in the spectrum of a gas heated to incandescence artificially in the laboratory. This remarkable observation was made by Young (q.v. ) at the Spanish eclipse of 1870, and repeated photographically in 1896 by Shackleton, at the Nova Zembla eclipse. The so-called chromosphere is a great layer composed of very hot gases, principally hydro- gen, and surrounding the .siui chiefiy outside the reversing laj'er. It is supposed, however, that both tliese strata are mixed together, except that the hottest and densest gases are in the lower or reversing layer. This chromospheric layer can be seen for an instant at the begin- ning or end of a total solar eclipse, and it is found to consist principally of a mass of red flames. The color is due to the preponderating presence of hydrogen, and it is this color that gives rise to the name chromosphere or color- sphere. The presence of flame does not. however, indicate a process of combustion in the usual sense of the word. For instance, coal burning in an ordinary stove is really being combined chemically with the oxygen of the air, and chem- ical combination of that character is certainly not going on in the sun. The most interesting features of the chromo- sphere are the prominences. These are great eruptive jets of red hydrogen flame bursting out- ward from the solar surface, sometimes to a dis- tance of hundreds of thousands of miles into space. They were first seen during total solar eclipses, when the briglit light of the central solar disk had been obscured b.y the interposed moon. I'nder ordinary circumstances the central solar light illuminates our terrestrial atmosphere very strongly, and enough of this light is re- flected from the air itself into our telescopes to mask completely the fainter light of the promi- nences. But it has been found possible to study the prominences sppctroscopically during full sunlight, and without waiting for a solar eclipse. This was first done visually in 18(18 by Janssen, Lockyer (q.v.), and Iluggins (q.v.). and photo- graphically by Hale (q.v.) and Deslandres in I8'J0. The two latter astronomers working inde- pendently of each other, the former in Chicago and the latter in Paris, devised certain instru- ments called photographic spectro-heliographs, by means of which it is possible to take pictures of the chromosplicre and prominences over the entire solar surface at once and in full sun- light. It has thus become known that the veloc- ity of motion in these prominence eruptions may be as great as several hundred miles per second, indicating real explosions of incalculable force. Kxtending far out beyond the chromosphere and prominences, and usually recognized as the sun's outermost envelope, is the corona. It is visible only at the time of total eclipse, when the central part of the solar disk is completely obscured. Tlie ajipearaiice of the corona at such times is very sudiicn ; tlie advancing edge of the moon slowly and gradually covers mure and more of the sun's surface, until finally only a sickle- shaped filament remains. Tlien suddenly this too disappears, and on the instant the superb corona bursts into view, truly one of the most magnificent objects in the whole range of nature's phenomena. 'The inner corcma, close to the solar disk, is very bright, even dazzling; farther out it fades bj' insensible gradations into a beauti- ful crown of filmy light. There are thread-like streamers, interlacing with complex involutions,' and extending often many millions of miles into space above the solar surface. Although observed visually from the most an- cient times, it was not until the application of photography to astronomical observation that we have been able to fix with some approach to certaintj- the details of the coronal structure. See ASTHO-PlIOTOGR.PIIT. Specti'oscopically the corona reveals the pres- ence of luminous gases, possibly containing in suspension particles of solid or liquid matter. No existing theory will explain quite satisfac- torily its mechanical construction, so difficult is it to reconcile with griivitational law so enormous an extension outward from the sun. Possibly magnetic forces are wholly or partly responsible for its existence and equilibrium; certain it is that we still have in the corona one of the 'pending problems' of astronomy. The chemical constitution of the sun has been studied .with great care spectroscopically. The spectroscope, when suitably arranged for the purpose, enables us to compare solar light with light obtained in the laboratory from vapors of terrestrial substances raised to incandescence by artificial heating. In this waj- many ter- I'estrial chemical elements have been identified in the sun ; and we may consider extremely probable the hypothesis that sun and earth are composed of the same kinds of matter. This, if true, is a fact of great importance in its bearing on the origin and evolution of the solar system. ( See Nebul.is : Ccsmogony.) The most conspicuous phenomenon of the solar surface is the snn-spots. These are apparently great holes or depressions in the photospheric surface, usu- ally having a dark spot in the middle, sur- rounded by a sort of radial grating. They vary in diameter from 500 to 60.000 miles for the central blaek parts, and range up to 1.50.000 miles for the surrounding less dusl^y portions. At times they are large enough to be seen with the naked eye. Their depth is very uncertain, and has been variously estimated from a few miles up to 2.500. The faenUr are bright streaks seen in various parts of the photospheric surface, but usually near spots, and most conspicuous when approaching the edge of the solar disk. The spots usually begin as mere dots, and then grow gradually or rapidly to the enormous di- mensions already stated. At times they break into several pieces, giving rise to a group or
