stick into a vessel of water: the stick will appear bent at the point where it enters the liquid, as in Fig. 4, thus giving the appearance as if the stick were lifted or bent upwards. Another very instructive experiment is to place a coin at the bottom of a vessel, and then standing so that the coin is just hidden by its edge, to gradually fill the vessel with water, the coin will appear to rise with the bottom of the vessel, and will become visible, as shown in Fig. 5.

The amount of refraction varies with the medium employed, and also with its temperature. The effect of

FIG. 5.-Refraction of light. Apparent elevation of the bottoms of vessels. different media can be clearly seen by passing a ray of light into a vessel, containing a liquid such as bisulphide of carbon, with a layer of water floating on the top. The ray will be seen to be bent on entering the water, and still more bent on passing from the water into the layer of bisulphide of carbon.

We have now to see what takes place when a ray of light enters a piece of glass. We will take first the case of glass with parallel sides. The ray on entering the glass at the upper surface is refracted downwards, as in the case of water, and travels through the glass

until it reaches the under surface. Here we have precisely the reverse condition holding,-that is, the ray of light passes from a dense medium to a rarer one; the ray is refracted upwards or away from the perpendicular line, and thus will exactly neutralize the previous refraction, and the beam of light will come out in a direction parallel to its original path, though not quite in the same straight line; as shown in Fig. 6, the ray instead of proceeding in the direction of s' proceeds in the direction of s.

FIG. 6.-Light passing through plate of glass.

If, then, a ray of light passes through a piece of glass, such for instance as a window glass, the surfaces of which are parallel, and inclined to the beam, you see when the beam passes through that the refractive effect is imperceptible. The reason of this is, that when we get the light falling on the glass from the air, then travelling through the glass, and coming into the air again under exactly the same. conditions, what is done at the first surface is exactly undone at the second, so that we get pretty much the same effect as at first. But now, if instead of having the glass bounded by parallel surfaces, we

use a wedge-shaped piece, or a prism, the sides of which are no longer parallel, you will see that there

FIG. 7.- Deviation of luminous rays by prisms.

is a distinct alteration in the effect produced; the beam is directed to another portion of the wall altogether. The ray strikes the first side of the

prism, and is bent towards the thicker part, or towards a line perpendicular to this surface, and on reaching the second side of the wedge, the ray is

Fr. 8.-Images of objects seen through prisms.

again bent in the same direction, towards the base of the prism; for in this case the ray is bent away from the perpendicular to the second surface, and the light emerges from the second surface in a totally

new direction. Fig. 7 shows the effect in three cases: the incident ray SI, the path in prism IE, and the refracted ray ER; NI and EN' being the lines

An experiment may

easily be tried, which will confirm this. Let a triangular piece of glass be held, with a refracting edge at top,

FIG. 9-Decomposition of light by the prism. Unequal refrangibility of the colours of the spectrum.