is clearly brought out when the sun is observed near sunset,

for then the rays have a vast depth of air to traverse, and, accordingly, we find the atmospheric bands to be defined with increased vividness.

We have hitherto spoken of the light from the sun, without attempting to distinguish the light from the different parts of the sun. The spectroscope can, however, be applied to each part of the sun separately, and thus it has been possible to establish remarkable differences between the spectra of the several parts. The most interesting discovery made in connection with this subject was that accomplished almost simultaneously by Janssen and Lockyer, which has given a wide extension to the possibilities of solar observation. In our chapter on the sun we have dwelt on the very remarkable objects known as the solar prominences, which become visible on the rare occasions of a total eclipse of the sun. The remarkable discovery just referred to disclosed a method by which the solar prominences could be observed at any time with the aid of the spectroscope.

The prominences are composed of glowing gas, principally hydrogen; and the light, instead of being spread all over the spectrum, like ordinary solar light, is mainly concentrated into the few lines which constitute the spectrum of hydrogen. When the spectroscope is directed around the edge of the sun the rays of ordinary diffused sunlight are distributed over the whole length of the spectrum, and are, of course, correspondingly weakened. The light from the prominences, being all concentrated into the lines, is, therefore, relatively increased in intensity to such an extent that they become visible.

The first line of Plate XIII. represents the solar spectrum, on which a few only of the principal lines are introduced for comparison. It is, however, to be understood that the solar spectrum really exhibits thousands of lines when viewed with a spectroscope of adequate power. Among the varied spectra of the stars different types have been recognised, the characteristic features of which are also depicted. The first type of stars are those which are intensely white, of which we may take Sirius and Vega as examples. In the

spectra of these stars there are comparatively few lines. The lines of hydrogen are however very strong, but the metallic lines of sodium and magnesium are barely perceptible. The presence of hydrogen in the atmosphere of these stars has been confirmed in a very remarkable manner by the photographs of the spectra of these stars which have been obtained by Dr: Huggins. In these photographs, lines can be seen in that part of the spectrum invisible to the eye, and these lines also have been found to coincide with lines in the photographic spectrum of hydrogen.

The second type of stars exhibit spectra closely resembling that of the sun; a good illustration of this type is found in the star Capella. Aldebaran has also a spectrum belonging to the second type, though verging on the third. The chief hydrogen lines are still conspicuous, but many metallic lines are coming strongly out.

The spectrum of a Orionis, or Betelgueze, exhibits a typical spectrum of the third class, which contains many metallic lines and shaded bands; but the most characteristic feature of this type is the absence of the hydrogen lines.

Another most remarkable achievement in spectrum analysis has been accomplished by Dr. Huggins in his very beautiful discovery of the spectroscopic method of measuring the motion of stars along the line of sight. The place of a line of hydrogen, for instance, will be slightly moved to one side or the other, according as the star be approaching the earth or receding from the earth. The amount of displacement is very small; but in the case of several stars large enough to enable an approximation to be made to the velocity with which the star is moving towards or moving from the earth.

The theory of this method is beautifully verified by observations on the sun. As the eastern edge of the sun is approaching and the western is receding, there is a corresponding difference in the spectra of the two edges, and the observed amount gives a velocity of rotation practically coincident with that otherwise known.

CHAPTER XXIII.

STAR CLUSTERS AND NEBULE.

Interesting Sidereal Objects-Stars not Scattered uniformly-Star Clusters-Their Varieties-The Cluster in Perseus-The Globular Cluster in Hercules-The Milky Way-A Cluster of Minute Stars-Nebula distinct from Clouds— Number of known Nebula-The Constellation of Orion-The Position of the Great Nebula-The Wonderful Star @ Orionis-The Drawing of the Great Nebula in Lord Rosse's Telescope-Photographs of this Wonderful Object— The Importance of Accurate Drawings-The Great Survey-Photographs of the Heavens-Magnitude of the Nebula-The Question as to the Nature of a Nebula-Is it composed of Stars or of Gas?-How Gas can be made to Glow -Spectroscopic Examination of the Nebula-The Great Nebula in Andromeda. -Its Examination by the Spectroscope-The Annular Nebula in LyraResemblance to Vortex Rings-Planetary Nebula-Drawings of Several Remarkable Nebula-Distance of Nebula-Conclusion.

WE have already mentioned Saturn as one of the most glorious telescopic spectacles in the heavens. Setting aside the obvious claims of the sun and of the moon, there are, perhaps, two other objects visible from these latitudes which rival Saturn in the splendour and the interest of their telescopic picture. One of these objects is the star cluster in Hercules; the other is the great nebula in Orion. We may take these objects as typical of the two great classes of bodies to be discussed in this chapter, under the head of Star Clusters and Nebulæ.

The stars that, to the number of several millions, bespangle the sky, are not scattered uniformly. We see that while some regions are comparatively bare of stars, others contain stars in profusion. Sometimes we have a small group, like the Pleiades; sometimes we have a stupendous region of the heavens strewn over with stars, as in the Milky Way. Such objects are called star clusters. We find every variety in the clusters; sometimes the stars are remarkable for their brilliancy, sometimes for their enormous numbers, and sometimes for the remarkable form in which they are grouped.