The test applied by the optician is much more exact, and also more easy. He points the instrument at a star, or at the reflection of the sun's rays from a small round piece of glass or a globule of quicksilver several hundred yards away, and ascertains whether the rays are all brought to a focus. This is not done by simply looking at the star, but by alternately pushing the eye-piece in beyond the point of distinct vision and drawing it out past the point. In this way the image of the star will appear, not as a point, but as a round disk of light. If the telescope is perfect, this disk will appear round and of uniform brightness in either position of the eye-piece. But if there is any spherical aberration or differences of density in different parts of the glass, the image will appear distorted in various ways. If the spherical aberration is not correct, the outer rim of the disk will be brighter than the centre when the eye-piece is pushed in, and the centre will be the brighter when it is drawn out. If the curves of the glass are not even all around, the image will appear oval in one or the other position. If there are large veins of unequal density, wings or notches will be seen on the image. If the atmosphere is steady, the image, when the eye-piece is pushed in, will be formed of a great number of minute rings of light. If the glass is good, these rings will be round, unbroken, and equally bright. We present several figures showing how these spectral images, as they are sometimes called, will appear; first, when the eye-piece is pushed in, and secondly, when it is drawn out, with telescopes of different qualities. We have thus far spoken only of the refracting telescope, because it is the kind with which an observer would naturally seek to supply himself. At the same time there is little doubt that the construction of a reflector of moderate size is easier than that of a corresponding refractor. The essential part of the reflector is a slightly concave mirror of any metal which will bear a high polish. This mirror may be ground and polished in the same way as a lens, only the tool must be convex. [Illustration with caption: SPECTRAL IMAGES OF STARS; THE UPPER LINE SHOWING HOW THEY APPEAR WITH THE EYE-PIECE PUSHED IN, THE LOWER WITH THE EYE-PIECE DRAWN OUT. A The telescope is all right B Spherical aberration shown by the light and dark centre C The objective is not spherical but elliptical D The glass not uniform--a very bad and incurable case E One side of the objective nearer than the other. Adjust it]