ural, it is necessary, as already explained, to reverse the mounting. In Fig. 132 the mounting is thus reversed, as seen by the fact that points in the foreground, a a, are farther apart than in the background, bb. The

usual mode of representation is shown in this figure. The true visual result is shown in Figs. 133 and 134, of which Fig. 133 represents the result when the observer is regarding the background, and Fig. 134 when he is regarding the foreground. It is seen that not

only does the diagram give truly the place and distance of the combined image, but also of the double images by means of which perspective is perceived.

It will be remembered that double images may be nearer or farther off than the point of sight, but that in the former case they are heteronymous, in the latter homonymous. In this way we at once perceive their distance in relation to point of sight. Now, in the new mode of representation, this fact is also indicated. In both of the figures 133 and 134 there are two places where the ray-lines cut the median lines, and therefore where double images may be formed; but in the one case the images are heteronymous, and therefore we refer them to the rearer points a a'; in the other case they are homonymous, and therefore we refer them to the farther points b b'.

If stereoscopic pictures mounted in the usual way be combined with the naked eyes by squinting, or pictures with reverse mounting be combined in the stereoscope, the perspective will be inverted. In this case the diagrammatic representation is exactly the same, except that the double images of points in the foreground a a' will now be homonymous, and therefore referred to the other possible point of reference, viz., beyond the point of sight; and double images of points in the background bb' will become heteronymous, and therefore referred to the nearer point.

Some curious Phenomena illustrating the heteronymous Shifting of the two Fields of View.

Experiment 1.-To trace a picture where it is not. Take a postage stamp, or a piece of coin, or a medallion, or a small object or picture of any kind; place it on a sheet of white paper. Take then a thin opaque screen,

like a pamphlet, or thin book, or piece of cardboard, and set it upright on the right side of the object or picture, and bring down the face upon the top edge of the screen, in such wise that the latter shall occupy the median plane. If we now gaze with the eyes parallel -i. e., on vacancy-the median card will double and become two parallel cards, and in the middle between them will be seen the object or picture. With a pencil in the right hand we may now trace the outline of the object or picture, by means of its image, on the right side of the screen, although the actual object or picture is on the left side of the same.

The accompanying diagrams illustrate and explain the phenomena. In Fig. 135, R and L are the two

eyes looking down on the paper sheet sh; ms is the median screen, and c the coin on its left side; a, the spot where the outline is traced with the pencil P. This figure therefore gives the actual condition of things. The visual result, and therefore the explanation, is given in Fig. 136. By careful inspection it is seen that the screen is doubled heteronymously, and becomes two parallel screens ms, m's; that the two images of the

paper sheet are slidden over each other, so that the left eye, its visual line, and its image of the coin c are all brought to the middle, while the right eye, its visual line, and its image of the pencil and of the point a are also brought to the middle from the other side, and superposed. We therefore see the image of the coin and trace its outline exactly an interocular space distant from its real position. If it were not for the screen, there would be another (right-eye) image of the coin and another (left-eye) image of the pencil and of the point a. These I have indicated in dotted outline.

Experiment 2.-If we make the experiment without the use of the median screen, then the cause of the phenomenon becomes obvious. If we lay a piece of money on a sheet of paper, and then gaze in the direc tion of the coin, but with the eyes parallel-i. e., on vacancy-the money of course separates into two images an interocular space apart. If we approach this with a pencil for the purpose of tracing the outline, we will see the pencil also doubled. If we now bring corresponding images in contact-i. e., right-eye image (left in position) of the pencil with the right-eye image (left in position) of the coin-we touch the coin with the pencil. But if, on the contrary, we bring the right-eye image (left in position) of the pencil to the left-eye image (right in position) of the coin, we may trace the outlines of the piece an interocular space distant from its true position. This is shown in Fig. 137, which gives the visual result of such an experiment—c and c′ being the right- and left-eye images of the coin, and P and P' of the pencil. If, while the operation is going on, we observe carefully, we will see to the right the left-eye image of the pencil, P', engaged in making a tracing. But there is no tracing in this place; it is

only the left-eye image of the real tracing being made by the other pencil P. In the previous experiment the screen cuts off all the images except the right-eye image

FIG. 137.

of the pencil and the left-eye image of the coin, which are brought together in the middle.

Tolerably good tracings of a picture may be made in this way. The only difficulty in making them really accurate is the unsteadiness of the optic axes, and therefore of the place of the image. I have, however, used this method in making outline tracings of microscopic objects, which may be filled out afterward. For this purpose a card is placed on the right side of the microscope, and the microscopic object is viewed with the left eye, while the right eye is used for guiding the pencil. Precisely as in the experiment with the coin (Fig. 137), the left-eye image of the object and the right-eye image of the pencil and of a certain spot on the card are brought together in the middle.

Experiment 3.-To trace the outlines of a light on an opaque screen. The same experiment may be modified in an interesting way thus: Set a light in front of you on a table. Place a median screen of cardboard or of tin between the eyes, so that the light can be seen with both eyes. Now bend the screen to the right so as to make a right angle at the distance of 6 or 8 inches from the eyes. This part will cut off the view of the