The' importance of a detailed study of our nearest fixed star, the sun, has been empha sized by the erection by the Carnegie Institu tion at Washington of the Solar Observatory at Mount Wilson. The discoveries at this ob servatory have been epoch-making. The work at Mount Wilson has demanded (1) a large image of the sun; (2) that spectrographs be kept under as nearly constant conditions of temperature as possible. Hence there has been the erection by Hale of the 'tower telescope?' There was first erected a 60-foot tower, and this proving successful, there was later put up a tower 150 feet high. A beam of sunlight re flected from a plane mirror at the top of the tower passes through an objective and is brought to a focus on the slit plate of the spec trograph. From the slit, the beam passes ver tically down a 75-foot well, at the bottom of which is placed a concave grating. The splen did definition obtained in California, and the great dispersion of the large spectrograph, have permitted a detailed study of spots on the sun. The work of Hale has shown (1) that sun spots are cooler than the rest of the sun; (2) that the spots are vortices; (3) that they are centres of magnetic disturbance. These facts have been verified only after a great variety of experiments carried out at Mount Wilson and at the Solar Observatory Office in Pasa dena.
Eclipses have furnished interesting develop ments in the history of spectrum analysis. The spectroscope was first used at the eclipse of 1868 by Janssen in India, when it was shown that the prominences give a bright line spectrum, thus showing that they are masses of gas. The lines appeared so bright in the 'spectroscope that Janssen looked for them the next day without an eclipse and found them readily enough. In 1869 helium was discovered on the sun, though it was not found in the laboratory by Sir William Ramsay till 1895; in 1870 Young discovered the 'flash spectrum." Observations similar to Janssen's of 1868 were independently carried out by Lockyer in England, who found the prominences without the aid of an eclipse. Both communicated their results to the French Academy, and by a strange coincidence both papers were read at the same meeting of the Academy. Prominences thus ceased to be purely eclipse phenomena. Val uable observations were made by Young, who used visual methods. In 1890 Hale invented the spectroheliograph, by means of which it was possible to photograph the prominences. Quite independently, Deslandres of Paris in vented a similar instrument. The essential feature of the spectroheliograph is the second slit, which is placed directly in front of the photograph plate and is used to allow light of only one wave-length to reach the plate. Hale's spectroheliograph reached its highest develop ment on the 40-inch Yerkes telescope. By set ting the second slit on a certain line of the spectrum it was possible to photograph the prominence in the light of that line only. Pho tographs could thus be obtained in the light of glowing calcium vapor by utilizing the H or the K line of calcium, or from glowing hydro gen using the C or H a line. The exquisite de fining power of the 40-inch lens has permitted remarkable detail to be shown in the prom inence photographs. At times prominences reach the great elevations of 300,000 miles above the sun, and are shot up with velocities as great as 100 miles per second. The chief workers along these lines have been Hale, Des landres, Evershed, Fox, Slocum and Ellerman. They all have shown that prominences are in timately associated with sun spots.
In photographing the surface of the sun by means of the spectroheliograph, Hale discov ered bright patches of gas which he called "flocculi." In his investigations of flocculi, Hale found that the photographs showed differ ent appearances when the slit was shifted to a slight amount from the centre of the line of the spectrum which was being utilized. A photo graph may be taken with the slit of the spectro heliograph at the centre of the K line at X 3933.8, another with the slit moved a trifle to X 3932, and still others with the slit at A 3929 and X 3924. These photographs show bright "flocculi," but the four of them not only differ from photographs taken with the calcium H or with hydrogen C or F lines, but differ ma terially among themselves. These photographs
are explained as being due to a difference in level of the gases, and from these and other results it seems probable that the calcium flocculi are in general made up of a series of columns, which expand as they reach higher levels, and in many cases overhang laterally. Both prominences and flocculi may be photo graphed by a spectroheliograph attached to a horizontal telescope in which the light is fed to the objective by means of a plane mirror, as Hale has shown at Mount Wilson and Evershed in India. These horizontal telescopes, however, do not equal the definition obtained by the 40 inch telescope.
The dark Fraunhofer lines of the solar spectrum are caused by the absorption of light as it passes from the hot photosphere through the relatively cooler layers of atmosphere sur rounding the photosphere, the so-called "re versing layer." The gases forming the revers ing layer are cool only in contrast to the very hot matter making up the photosphere, but these gases, nevertheless, are at such a high temperature, that if we could obtain their spectrum separated from the photosphere, the spectrum would consist of a series of bright lines on a dark background. At the time of a total eclipse of the sun, we are permitted to obtain the spectrum of the reversing layer. So long as even a small portion of the sun's disc remains visible, the light from the photosphere is so overpowering compared with that of the reversing layer, that the ordinary, dark-line Fraunhofer spectrum is obtained. Just before totality a few bright lines of the upper chromo sphere become visible, first the H and K lines of calcium, and these are followed by the hydrogen and helium lines. At the instant that the sun's disc is entirely covered by the moon, there flashes out the bright line spectrum of the reversing layer. The change from dark-line to bright-line spectrum is so rapid that Young in 1870, who was the first to see it, called it the "flash spectrum." The flash remains visible in its entirety only for a few seconds, while the moon is passing over the relatively shallow re versing layer. The flash spectrum was first photographed in the year 1896 by Shackleton, who used a prismatic camera. The most com plete results so far published are those of Mitchell, who obtained exquisite definition on his photographs at the 1905 eclipse. The ap paratus used was extremely simple, and con sisted of a plane mirror to turn the sun's light horizontally. From the mirror, the beam of light fell on a concave grating by means of which the spectrum was brought to a focus on the photographic plate. No slit was used. Photographs were obtained extending from X 3300 in the violet to X 5900, with a dispersion of one mm equal 10.8 Angstroms. The wave lengths of 2841 lines in this region (Astrophys ical Journal, XXXVIII, 407, 1913), show an accuracy of 0.02 Angstroms. In addition to the wave-lengths, there were also given the heights to which the vapors forming each line of the spectrum extend in kilometres above the sun's photosphere. The general conclusions from this eclipse work are: (1) That the reversing layer is merely the densest part of the chromo sphere, and has no existence separate from the chromosphere; (2) that the reversing layer extends about 600 kilometres above the photo sphere; (3) that there is no difference in wave lengths between the spectrum of the reversing layer and Rowland's values for the Fraunhofer lines, showing that in reality the flash spectrum is truly a reversal spectrum ; but (4) there are great differences in intensities between the lines of the flash spectrum and the Fraunhofer spectrum. These intensity differences find a ready explanation in the differences in height to which the vapors extend. On account of the great heights, the H and K lines of calcium, the hydrogen series, and the helium lines are all strong in the flash spectrum. Very great dif ferences in intensity are shown by the enhanced lines.