the interposition of the moon at a total eclipse. But methods have been devised by which they can be observed or photographed on any clear day through the agency of a modified form of spectroscope. The prominences are constantly changing in form, sometimes slowly, as in the case of this group (Fig. 10), a photograph of which, taken at the eclipse of May 28, 1900, by the Astronomer Royal of England in Spain, is shown for comparison with the photograph taken about two hours earlier by the Yerkes Observatory party in North Carolina. Here the
Cloud-like Prominences photographed at the Eclipse of May 28, 1900. a, by Yerkes Observatory Party at Wadesboro, N. C. b, by Astronomer Royal of England at Ovar, Portugal, two hours later. (The bright cross on the right of this picture is due to a defect in the original photograph.)
change in the form of the mass of gas which constitutes the prominence, is comparatively small, but that violent forces are sometimes at work may be illustrated by photographs of an eruptive prominence taken at the Kenwood Observatory in 1895 (Fig. 11). At the moment when the first photograph was made the prominence had attained a height of 160,000 miles and was rising rapidly. Eighteen minutes later another picture was taken; during the interval the prominence had been going upward at the rate of six thousand miles a minute, and when the exposure was made it had reached an elevation of 280,000 miles. When looked for a few minutes later it had completely disappeared.
The constitution of the chromosphere, the sea of flame some 10,000 miles deep from which the prominences arise, increases in complexity as the surface of the solar disk is approached. In its upper part only the vapor of calcium and the light gases, hydrogen and helium, are found. But in proceeding downward the vapors of magnesium, sodium, iron, chromium, and last of all, carbon, are successively encountered. At this part of the solar atmosphere the dark lines of the solar spectrum take their rise through the effect of absorption.
Time does not permit a detailed description of the phenomena of