discovery referred to. If with a glass prism of a certain form we produce a spectrum of the sun, this spectrum will be thrown a certain distance away from the point on which the sun's rays would fall if not interfered with. This distance depends on the refractive power of the glass. The spectrum will have a certain length, depending on the dispersive power of the glass. Now, if we change our prism for another of exactly the same shape, but made of a different kind of glass, we shall find the spectrum thrown to a different spot. If it appeared that the length of the new spectrum was increased or diminished in exactly the same proportion as its distance from the line of the sun's direct light, it would have been hopeless to attempt to remedy chromatic aberration. Newton took it for granted that this was so. But the experiments of Hall and the Dollonds showed that there is no such strict proportionality between the dispersive and refractive powers of different kinds of glass. It accordingly becomes possible to correct the chromatic aberration of one glass by superadding that of another.

Fig. 4.

This is effected by combining, as shown in fig. 4, a convex lens of crown glass with a concave lens of flint glass, the convex lens being placed nearest to the object. A little colour still remains, but not enough to interfere seriously with the distinctness of the image.

But even if the image formed by the object-glass were perfect, yet this image, viewed through a single convex lens of short focus placed as in fig. 1, would appear curved, indistinct, coloured, and also distorted, because viewed by pencils of light which do not pass through the centre of the eye-glass. These effects can be diminished (but not entirely removed together) by using an eye-piece consisting of two lenses instead of a single eye-glass. The two forms of eye-piece most commonly employed are exhibited in figs. 5 and 6. Fig. 5 is Huyghens' eye-piece, called also the negative eye-piece, because a real image is formed behind the field-glass (the lens which lies nearest to the object-glass). Fig. 6 represents Ramsden's eye-piece, called also the positive eye-piece, because the real image formed by the object-glass lies in front of the field-glass.

Fig. 5.

Fig. 6.

The course of a slightly oblique pencil through either eye-piece is exhibited in the figures. The lenses are usually plano-convex, the convexities being turned towards the object-glass in the negative eye-piece, and towards each other in the positive eye-piece. Coddington has shown, however, that the best forms for the lenses of the negative eye-piece are those shown in fig. 5.

The negative eye-piece, being achromatic, is commonly employed in all observations requiring distinct vision only. But as it is clearly unfit for observations requiring micrometrical measurement, or reference to fixed lines at the focus of the object-glass, the positive eye-piece is used for these purposes.

For observing objects at great elevations the diagonal eye-tube is often convenient. Its construction is shown in fig. 7. ABC is a totally reflecting prism of glass. The rays from the object-glass fall on the face AB, are totally reflected on the face BC, and emerge through the face AC. In using this eye-piece, it must be remembered that it lengthens the sliding eye-tube, which must therefore be thrust further in, or the object will not be seen in focus. There is an arrangement by which the change of direction is made to take place between the two glasses of the eye-piece. With this arrangement (known as the diagonal eye-piece) no adjustment of the eye-tube is required. However, for amateurs' telescopes the more convenient arrangement is the diagonal eye-tube, since it enables the observer to apply any eye-piece he chooses, just as with the simple sliding eye-tube.

Fig. 7.

We come next to the important question of the mounting of our telescope.

The best known, and, in some respects, the simplest method of mounting a telescope for general observation is that known as the altitude-and-azimuth mounting. In this method the telescope is pointed towards an object by two motions,—one giving the tube the required altitude (or elevation), the other giving it the required azimuth (or direction as respects the compass points).

For small alt-azimuths the ordinary pillar-and-claw stand is sufficiently steady. For larger instruments other arrangements are needed, both to give the telescope steadiness, and to supply slow movements in altitude and azimuth. The student will find no difficulty in understanding the arrangement of sliding-tubes and rack-work commonly adopted. This arrangement seems to me to be in many respects defective, however. The slow movement in altitude is not uniform, but varies in effect according to the elevation of the object observed. It is also limited in range; and quite a little series of operations has to be gone through when it is required to direct the telescope towards a new quarter of the heavens. However expert the observer may become by practice in effecting these operations, they necessarily take up some time (performed as they must be in the dark, or by the light of a small lantern), and during this time it often happens that a favourable opportunity for observation is lost.

These disadvantages are obviated when the telescope is mounted in such a manner as is exhibited in