mixing with it a few drops of milk, and the room be filled with dust by sweeping a carpet, or any other means, then you will see the refraction which the ray from the shutter undergoes in passing into the water, the reflection of it at the surface of the looking-glass, and the refraction which takes place when the ray leaves the water, and passes again into the air.

James. Does this refraction take place in all kinds of glass?

Tutor. It does; but where the glass is very thin, as in window glass, the deviation is so small as to be generally overlooked. You may now understand why the oar in the water appears bent, though it be really straight; for suppose A B (Plate 1.

Fig. 4.) represent water, and m a r the oar, the image of the part a r in the water will lie above the object, so that the oar will appear in the shape man, instead of max. On this account also, a fish in the water appears nearer the surface than it actually is, and a marksman shooting at it must aim below the place which it seems to occupy.

Charles. Does the image of the object seen in the water always appear higher than the object really is?

Tutor. It appears one fourth nearer the surface than the object is. Hence a pond or river is a third part deeper than it appears to be, which is of importance to remember, for many a school-boy has lost his life

by imagining the water into which he plunged was within his depth.

James. You say the bottom appears one fourth nearer the surface than it and then that the water is a third

is; deeper than it seems to be: I do not understand this.

Tutor. Suppose the river to be six feet deep, which is sufficient to drown you or me, if we cannot swim: I say the bottom will appear to be only four feet and a half from the surface, in which case you could. stand and have the greater part of your head above it; of course it appears to be a foot and a half shallower than it is; but a foot and a half is just the third part of four feet and a half.

Charles. Can this be shown by experiment?

Tutor. It may :-I take this large empty pan, and with a piece of soft wax stick a piece of money at the bottom, but so that you can just see it as you stand; keep your position, and I will pour in a quantity of water gradually, and tell me the appearance.

Charles. The shilling rises exactly in the same proportion as you pour ia

the water.

Tutor. Recollect then, in future, that we cannot judge of distances so well in water as in air.

James. And I am sure we cannot of magnitudes: for in looking through the sides of a globular glass at some gold and silver fish, I thought them very large; but if I looked down upon them from the top, they appear much smaller indeed.

Tutor.

Here the convex or round shape of the glass becomes a magnifier, the reason of which will be explained hereafter. A fish will, however, look larger in water than it really is. I will show you another experiment which depends on refraction here is a glass goblet twothirds full of water; I throw into it a shilling, and place a plate on the top of it, and turn it quickly over, that the water may not escape. What do you see?

Charles. There is certainly a half crown lying on the plate, and a shilling seems swimming above it in the

water.

Tutor. So it appears indeed; but it is a deception which arises from your seeing the piece of money in