One of the most successfully observed of recent eclipses was that of May 28, 1900. The duration of totality was only two minutes, but almost perfect weather prevailed everywhere. It was visited by a large number of skilled observers, and an examination of the work performed and attempted will give a good idea of what astronomers at the present day hope to learn about the sun at times of total eclipse. As stated above, the ordinary solar spectrum consists of a bright band crossed by dark absorption lines due to a reversing layer present in the chromosphere. At the eclipse of 1870, Professor C. A. Young, who was watching the spectrum of the fast disappearing sun, saw, at the instant when the last bit of the photosphere was covered by the moon, the solar spectrum with its dark lines replaced by a spectrum composed of bright lines. This phenomenon, from the suddenness of its appearance became known as the 'Flash.' The 'flash' spectrum is one of the most interesting features of a total eclipse. The depth of the flash layer is very small, and the duration of its greatest intensity very brief, since it is covered by the moon after two or three seconds. To obtain good photographs of this phenomenon is somewhat difficult. This has been accomplished, however, at the eclipses of 1896 and 1898, and, especially, by several observers, at the eclipse of 1900. Several kinds of spectroscopes are in use. Ordinarily an astronomical spectroscope consists of a telescope, a narrow slit, a train of prisms, and a small telescope which brings the spectrum to the eye or to the photographic plate. When the object which is to be examined has an area like the sun the use of a slit cannot be avoided. When the source of light is a point, or a narrow line of light, there is no such necessity and the more simple apparatus, known as the slitless spectroscope, or objective prism, may be used. This consists of a prism placed over the lens of the telescope and a photographic plate at the focus. Instead of the prism or prisms a diffraction grating may be used. Professor Pickering, the director of the Harvard Observatory, has obtained for many years fine spectra of the stars by this method, which is an adaptation of the original method of Fraunhofer. An apparatus of this sort used in eclipse work is known as a 'prismatic camera.' It is evident that this form of spectroscope could not be successfully used on the uneclipsed sun, since the resulting spectrum would be simply a confused mass of colored light. There must be a slit, but in the case of total eclipse, nature furnishes it. As the moon at such times has an apparent diameter greater than that of the sun, it is readily seen that at the instant before the moon's disc completely covers the sun there will remain a very narrow crescent of light. At the instant after totality has begun the photosphere will be entirely covered, but for two or three seconds the thin line of chromospheric light remains in view. The two spectra taken at these moments, the one an instant before
