that all the light that falls on D previously fell on C. E yet remains in darkness. Next, remove D and replace C, and E still remains in shade; but, on removing C, E is fully illuminated. This shows that the quantity of light that fell on C spreads over four times the surface at the distance D, and nine times the surface at the distance E. Hence each one of the squares on D is one-fourth as bright as the square C, and any one of the squares on E is one-ninth as bright as C.

Here we are coming upon another fact about light; we find another law governing its action. At one foot from the lamp the light had a certain power; at two feet it had only one-quarter as much power; at three feet it only had one-ninth as much power or intensity. So, if we approach the lamp, at a certain distance the light has a certain brightness; at half that distance it has four times the brightness; at one-third the distance it has nine times as much brightness. The above relation, existing between the intensity of the light on a surface at different distances from the source of light, is often stated as follows: The illumination of a given surface varies in brightness inversely as the square of its distance from the source of light.

EXPERIMENT IN MEASURING LIGHT.

This picture represents a sheet of white paper, standing upright upon a table. A few inches from this screen is our needle-pointed awl, stuck upright in the table. (If you do not care to do this, the awl can be stuck into a block of wood or bit of wax.) A lighted candle is placed on the table, about 22 inches (55.8 centimetres) from the screen. Beyond this is a lamp, placed upon a pile of books, so as to bring the

FIG. 7.

flame of the lamp on a level with the flame of the candle. The lamp should stand, say, at 44 inches (112 centimetres) from the screen; and, if it has a flat wick, it must be so placed that the wick stands diagonally to the screen.

These are all the things we need to make a most

interesting experiment in measuring light, and we have only to make the room dark, or put out all the other lights, if it is evening, and we can go on with the work. Let the candle burn a moment or two, and then bend the wick down, so as to give a large flame. If you have no lamp, a gaslight will answer. Upon the paper screen are two shadows of the awl side by side. Move the lamp to the right or left till the two shadows just touch, and make one broad band. Study this double shadow carefully. Perhaps one half is darker than the other. Move the lamp backward or forward, and you will see that its shadow changes becomes darker or lighter. Presently you will find a place for the lamp where the double shadow appears of a uniform depth.

Now both lamp and candle cast just as deep a shadow, and yet one is much farther from the screen than the other. Measure off the distance. Perhaps the candle is 22 inches (55.8 centimetres) from the the screen, and the lamp is 44 inches (112 centimetres).

In our last experiment we found that the illumination of a given surface varies inversely as the square of its distance from the source of light. The square of 22 is 484, and the square of 44 is 1,938. Now, if we divide 1,938 by 484, we get 4, and thus

we find that our lamp is four times as bright as the candle. It casts just the same depth of shadow on the screen as the candle, and it is four times as bright, because the square of the distance of the candle will divide the square of the distance of the lamp four times. If we measure it another way we find the candle is half the distance from the lamp to the screen, and gives only one-quarter as much light.

Such a measurement as this is both easy and simple, and by means of such an experiment we can find out how much light any lamp gives. In this case, we find one lamp gives just four times as much light as the candle, or as much light as four candles would give at once. This is called a photometric experiment, from two Greek words meaning lightmeasurement. You may sometimes hear people say that a certain gas-lamp gives a sixteen or eighteen candle light, and our experiment shows us what they mean by this expression. They mean that the lamp has a photometric value of so many candles, or gives a light equal to the light of sixteen or eighteen candles burning at the same time.

CHAPTER III.

REFLECTION OF LIGHT.

PLACE the heliostat in position, and bring a slender beam of light into the darkened room. Then get a small looking-glass, or hand-mirror, and a carpenter's steel square, or a sheet of stiff paper, having perfectly square corners. Hold the mirror in the beam of light. At once you see there are two beams of sunlight, one from the heliostat and another from the mirror. Hold the glass toward the heliostat, and you will see this second beam going back toward the window.

This is certainly a curious matter. Our beam of light enters the room, strikes the mirror, and then we appear to have another. It is the same beam, thrown back from the glass. This turning back of a beam of light we call the reflection of light.

Place a table opposite the heliostat, and place the mirror upon it, against some books. Turn the mirror to the right, and the second or reflected beam of light moves round to the right. Turn the glass still more, and the beam of light will turn off at a right angle, and there will be a spot of light on the wall at that