prism of water, it is turned in the direction CD. If you examine the prism, (as it is usually called in Optics, meaning the same form as that of a trough), you find that when the point of it is downwards, the effect of it is that the beam of light which comes in this direction AB, is turned in the direction CD, or more upwards. There is a rule on this matter, which is thus expressed that the course of the light is always turned to the thicker part of the prism. Or if you observe what is the bending of the light at the two surfaces of the prism, this is the way in which it may be expressed-when the light comes from the air into the water, its direction is bent more nearly towards the direction of the line which is perpendicular to the surface—when it goes from the water to the air, it is bent further from the perpendicular.*

In the

particular use of the prism, with its point downwards, these two things are combined in such a manner, that at each of these surfaces the direction of the beam of light is bent upwards. Of course you will infer that if the prism were turned in the oppo

D

FIG. 5.

site way as at Figure 5, so that its point was upwards,

*The reader is particularly desired to remark that the word perpendicular does not mean perpendicular to the horizon, or vertical, unless it is so expressed. When the expression perpendicular to the surface of the glass is used, it means what a workman would probably call square to the surface of the glass. The vertical direction at any place is that of a plumbline hanging there, or perpendicular to the surface of still

water.

then the course of the light would be bent downwards.

Now, as regards astronomical observations, we have no water or glass concerned; but we have a thing which produces refraction, and that is atmospheric air. The common air produces refraction. The visible exhibition of this refraction is one of those nice experiments which I cannot attempt to exhibit to an audience like this. But it may be shown in various ways; as, for instance, by forming a prism of glass, and compressing more air into it; or again, by exhausting the air from it. It is shown that the effect of air is precisely the same in kind as the effect of water, though much less in degree. It may be stated as a general law, that where light enters from external space into air, or into water, or glass, or diamond, if you please, or any other transparent substance-where light enters from external space into any one of these substances, its course is bent in such a direction that it is more nearly perpendicular to the dividing surface than it was before. Now, having laid that down as a general law, let us see what its application will be to atmospheric air. In making astronomical observations,

earth, covered by atmosphere-the black part being the earth, the dusky part the atmosphere. Suppose a beam of light is coming in the direction AB from a star, and suppose that at B it comes on the atmosphere it is coming then here exactly under the same circumstances in which in Figure 5 the beam of light comes upon the surface of the prism. According to the law which I have just mentioned, it will be bent in such a manner that its direction after it has entered the atmosphere is more nearly perpendicular to the bounding surface than before. Therefore, in conformity to that law, it is bent in the direction BC, and it reaches the eye of the observer at C, in the direction BC. If you observe the relation which the second line BC has to the first AB, you will see it is more nearly perpendicular to the horizon; or, standing at C on the surface of the earth, you have to look a little higher to see the star than if you were on the outside of the atmosphere, at B; or the star, in consequence of the action of the atmosphere, appears higher. Now this I have mentioned would be the case if the atmosphere had a definite boundary, and were uniform throughout its extent; but the same thing takes place if the atmosphere has not a definite boundary, and varies in density from stratum to stratum—the same effect takes place from one stratum of the atmosphere to the next.*

In this manner we find there is a rational explanation of this too great elevation of the stars. Taking as foundation the established law of optics, determined by experiments on glass and water, and computing from this what ought to be the deflection of light, and what ought to be the elevation of the star produced by the refraction of light by the atmosphere, and applying that as a correction to the observations made by the Equatoreal Instrument, of which I have spoken, it is proved that the whole thing comes quite right—that the stars move exactly in circles, not approximately, but (as far as the human eye and instruments can discover) exactly as if they turned uniformly round one imaginay axis. This is the grand fact which must be regarded as the foundation of Astronomy.

I shall now mention, in as few words as I can, how observations of all kinds are made, and how upon these observations the most accurate astronomical determinations are based. In the first place we will show the use of the telescope, and how it is used with wires in the field of view. The instrument thus fitted up is not used for mere gazing, but for accurate observation. If you go into an observatory, and look into any of the telescopes, you will see a set of bars. It will be perhaps beyond your comprehension what these bars are, and what they are for. Stars are seen to pass these

BD. Again, when it reaches cE, the boundary of the next stratum, it will in like manner be bent in the direction cd. The same thing will happen every time it comes to the boundary of a new stratum; and at last, when it reaches the earth's surface at C, its direction will be Cg. The star, instead of appearing to the observer at A, will consequently be seen at I, in the direction of Cg.

as if the stars and the bars were at the same distance from the eye. These bars are in reality fine cobweb threads, or something of the kind, fixed in the telescope very near to the eye. Perhaps Figure 7

may serve to illustrate the construction of the telescope. There is no tube, but that is immaterial. At A is what we call a lens, that is to say, a piece of glass convex on both sides, and therefore thickest in the middle. It is here supposed to be fixed in a hole in a wooden screen MN. The property of this lens of glass is, if there be a luminous object in the distance, it collects all the light from that object; and instead of suffering it to go out in a broad sheet of light, it makes it contract so that the light from each point in the object is collected at a corresponding point on the screen; and therefore all the corresponding points of light on the screen, which belong to the original points of light in the original luminous appearance, when put together, form an image which is exactly similar to the original object. The image, however, is turned upside down, because the light which comes from the upper part of the luminous object and goes through the lens, passes downwards towards the lower part of the screen KL.