first sight this might seem to indicate a contrary rotation of the eyes, viz., inward. But not so; for, observe, the field of view of the right eye is the left or black square, and the field of view of the left eye is the right or

FIG. 62.

B

white square. The right-eye field turns to the left, showing a rotation of the right eye to the right; while the left-eye field turns to the right, showing a rotation of the left eye to the left. Thus the two eyes in convergence rotate outward. This is shown in the diagram Fig. 62, in which cc' is the experimental plane. The arrows show the direction of rotation of the images of the plane and of the eyes.

Experiment 2.-When one becomes accustomed to experiments of this kind, he can make them in many ways. I find the following, one of the easiest and most convenient: Measure the exact height of the root of the nose upon the sash of the open window, and mark it. Stand with head erect about 3 or 4 feet from the window. Using the cross-bars of the sash-frame as horizontal lines, arrange the head so that the two images of the root of the nose shall be exactly the same height as the mark. The primary plane is now horizontal. Now converge the eyes until the dark outer jambs or sides of the frame of the sash approach each other. This will be very distinct on account of the bright light between them. It will be seen that the frames come

together, not parallel, but as a sharp V, thus

7 being the right- and left-eye images respectively. I find that when I stand at a distance from the window equal to the width of the sash, the angle between the two jambs as they come together is about 15°, showing a rotation of each eye outward 7° 30'. When standing still nearer, so that the convergence is extreme, the angle is 20° or more, showing a rotation of each eye of 10° or more.

In all these experiments the extremest care is necessary to insure the perfect horizontality of the visual plane. The slightest upward or downward looking vitiates the result by introducing mathematical perspective. If there were no rotation, then looking upward and converging would bring the jambs together by

perspective, thus— A; looking downward, thus

looking horizontal, parallel, thus-. But on account of rotation, looking horizontal brings them together

thus—"\/'; downward, at higher angle, thus—"\/\/\".

Looking upward more and more, the angle decreases till it becomes 0 (i. e., the jambs parallel), and then inverted. I find that in the previous experiment, standing from the window the distance of its width, I must elevate the plane of vision about 6o-i. e., I must look about 8 or 9 inches above the mark-to make the jambs parallel. This is therefore a good method of measuring amount of rotation.

Experiment 3.-A far more accurate mode of measuring the amount of rotation is by constructing diagrams on a plane similar to the one used in experiment 1, but in which the verticals and horizontals are both inclined on the true verticals and true horizontals in a

direction contrary to the rotation of the eyes-i. e., inward-and then determining the degree of convergence necessary to make them come together perfectly parallel. I find that for my eyes, when the verticals are thus inclined in each square 11° with the true vertical, and therefore make an angle of 21° with each other (Fig. 63), they come together parallel when the point of sight is 7 inches from the root of the nose. When the angle of inclination in each is 24° with the true vertical, and therefore 5° with each other, the point of

VERTICALS AND HORIZONTALS INCLINED 11°.

sight must be 4 inches off. When the inclination with the true vertical is 5°, and therefore 10° with each other, the point of sight is 2-2 inches. Finally, when the inclination with the true vertical is 10° or 20° with each other, then they can be brought together parallel only by the extremest convergence, the point of sight being then only a quarter of an inch in front of the root of the nose. In the diagram Fig. 63 the lines, both vertical and horizontal, are inclined inward 110, and therefore the verticals of the two squares make an

angle with each other of 21°. It is therefore a reduced facsimile of the plane used. The coördinate lines coincide when the point of sight is 7 inches from the root of the

nose.

In the cases of extreme convergence mentioned above, I find that for perfect coincidence of both verticals and horizontals it is necessary that the inclination of the verticals with the true vertical must be greater than that of the horizontals with the true horizontal; so that the little squares are not perfect squares. Thus, when

the verticals incline 5°, the horizontals must incline only 34°; when the verticals incline 10°, the horizontals incline only 5°. Fig. 64 is a reduced facsimile of this last case of extreme convergence. I can not account for this, except by a distortion of the ocular globe by the unusual and unnatural strain on the muscles, especially the oblique muscles of the eyes. It may be that other eyes are more rigid than mine, and suffer less distortion.

The above is by far the most refined method of proving rotation, and of measuring its amount. But

so difficult are these experiments, and so delusive the phenomena, that it is necessary to prove it in many ways. Another method is by means of ocular spectra. We have already shown that these are not so well adapted to experiments in convergent motion as they are in parallel motion. For example, two brands on the vertical meridians of the two retinæ produce spectral images which are perfectly united (p. 178). Now in strong convergence, when the two eyes rotate outward, the two images will not separate or cross each other, thus- , as we might at first expect; for this is forbidden by the law of corresponding points. But we may use a spectral image of one eye to show rotation of that eye.

X

Experiment 4.-The manner in which I conduct the experiment is as follows: I make a vertical spectral image in the manner already explained (page 164), by gazing with one eye (say the right) on a vertical slit in

a closed window. I now turn about, and, keeping the left eye L, Fig. 65, still shut, I look across the root of the nose n with the right eye R at a perfectly vertical line w on the wall. I see the vertical image perfectly parallel and nearly coincident with the vertical line on the wall. Then, while the right eye still continues to look along the line R s, I turn the shut left eye L from