image to the screen. These two sockets must, of course, be fitted to the apparatus, as it is necessary that the lengths B C (Fig. 6) should in this case be such as to bring the lens and reflector central with the perpendicular axis; but the ens and mirror already described will answer for the rest perfectly well. It might be supposed that the first surface of the mirror would cause a double image on the screen; but it is so faint in comparison with the other, that any such effect is rarely perceivable at all, and only then to close and special observation.

Besides its other more strictly experimental uses, the vertical attachment is very effective for projecting on the screen extempore diagrams, which are easily traced on a sheet of prepared glass laid horizontally. It often makes a demonstration much clearer to work out a somewhat complicated diagram step by step in this way. The ground glass and pencil is best adapted for this method of work, as the operator can work more freely and see more precisely what he is doing.

CHAPTER II.

RAYS AND IMAGES.-REFLECTION.

Rays of Light-Rays form Images-Inversion of Images-ShadowsLaw of Intensity at various distances-Law of Reflection-Virtual and Multiple Images-The Doubled Angle of Reflection—Application of this in the Reflecting Mirror-Reflection from Concave and Convex Mirrors-Images produced by Concave MirrorsScattered Reflection-Light Invisible.

11. Rays of Light.-All objects that are visible to us, are so in virtue of "rays of light" (whatever these may prove to be) which proceed from every point in the luminous surface. That these rays are perfectly straight or rectilinear while traversing the same medium, observation convinces us. We can trace the straight path of a sunbeam in a dusty room; we know, or find by experiment, that light from a lamp can only be seen through three perforated cards arranged with an interval between them, if all three apertures are in a perfectly straight line; and we are conscious that if any opaque body be held in the straight line between our eye and a lamp the light is effectually stopped. But the rest of the mechanism-the real image-forming power of these rays-is masked to the ordinary observer by the number of them. To see it clearly, we must isolate the rays which proceed from the object in certain directions.

12. Images. We may do this by darkening a room and

only allowing light to enter through a very small hole in the shutter the original camera obscura. Bringing a sheet of white cardboard near the aperture (Fig. 14), we see depicted

upon it the landscape outside, and we thus learn experimentally that rays of light are really sent out from all points of visible objects, and that these rays form images.

FIG 14-Image formed by a small hole.

This is equally the case whether the objects are self-luminous, like the sun or a lamp (with which class of bodies we more commonly associate the idea of this ray-sending property), or whether they only reflect to us light derived originally from other luminous sources. In all cases the image is formed on the retina or on a screen, by rays which proceed from the object to the site of the image.

I

This landscape image is rather faint, because the objects depicted are not very brilliant, and there are too few rays passing through the aperture to form a bright one. We can get a brighter image by collecting wider cones of rays from each point of the object, methods of doing which will speedily come before us; or the more brilliant light of the lantern will yield a better one. Remove all the lenses, including the condensers, and cover the flange-nozzle with a piece of tinfoil or a cap of black card. In this prick a hole with a rather thick needle, and an image of the radiant at once appears on the screen. Since each point of this image is defined by straight lines proceeding from the corresponding point of the radiant through the prick to the screen, it is obvious that if we prick another hole we must get another image. We go on pricking holes, an image appearing with each, till by degrees the opaque tinfoil or card is removed. As we do so the images of the radiant crowd on one another more and more; presently they touch;

2

1 A simple and striking experiment for the private student is to take off the top, and knock out the bottom of a coffee canister, and blacken it inside. Then punch a hole in the middle of its length, about one-sixteenth of an inch in diameter. Hold this over a naked candle in an otherwise dark room; and a good image of the candle-flame will be formed at six or eight inches' distance upon a white card or finely-ground glass.

2 A needle-prick will give good image from a lime-cylinder. With lower illuminants, the prick will have to be made larger, and a sheet of card, or the portable screen, should be brought within a few feet of the lantern.

then they overlap; at last, when all the tinfoil is gone, the screen is covered with a white glare of light.

13. Necessity for Isolating Phenomena.-We now see how the image-producing power of the rays proceeding from any luminous object is ordinarily masked by the mere multiplicity of such rays. We say the screen is "lighted," but we have found that this illumination really consists in its being covered over by an infinite number of images of the radiant in the lantern. We thus learn at the outset a lesson of the greatest importance, viz., that in many cases, to ascertain the true nature of the phenomena of Light, we must isolate a comparatively small number of rays. Unless we do so,

A

FIG. 15.-Inversion of image.

such a number of images (or other phenomena under examination) may get mixed up, or partially superposed, that their real character may be lost in a general confused body of light due to them all.

14. Inversion and Relative Size of Images.—It is plain, also, why such images as we have been forming must be inverted. Let o, Fig. 15, represent the original object; then it will be seen that the rays from its top and bottom, or right and left, if they proceed in straight lines, must cross at the aperture A, those from the top proceeding to form its image I at the bottom. It is further clear that the relative sizes of