Another form, analogous to this, but without the means of moving the pictures, is shewn in Fig. 16, as made by M. Duboscq. The adjustment is effected by moving the

FIG. 16.

eye-pieces in their respective tubes, and by means of a screw-nut, shewn above the eye-pieces, they can be adapted to eyes placed at different distances from one another. The advantage of this form, if it is an advantage, consists in allowing us to use larger pictures than can be admitted into the box-stereoscope of the usual size. A box-stereoscope, however, of the same size, would have the same property and other advantages not possessed by the open instrument.

Another form of the lenticular stereoscope, under the name of the cosmorama stereoscope, has been adopted by Mr. Knight. The box is rectangular instead of pyramidal, and the adjustment to distinct vision is made by pulling out or pushing in a part of the box, instead of the common and better method of moving each lens

separately. The illumination of the pictures is made in the same manner as in the French instrument, called the cosmorama, for exhibiting dissolving views. The lenses are large in surface, which, without any reason, is supposed to facilitate the view of the binocular pictures, and the instrument is supported in a horizontal position upon a stand. There is no contrivance for adjusting the distance of the lenses to the distance between the eyes, and owing to the quantity of light which gets into the interior of the box, the stereoscopic picture is injured by false reflections, and the sensibility of the eyes diminished. The exclusion of all light from the eyes, and of every other light from the picture but that which illuminates it, is essentially necessary to the perfection of stereoscopic vision.

When by means of any of these instruments we have succeeded in forming a single image of the two pictures, we have only, as I have already explained, placed the one picture above the other, in so far as the stereoscope is concerned. It is by the subsequent action of the two eyes that we obtain the desired relief. Were we to unite the two pictures when transparent, and take a copy of the combination by the best possible camera, the result would be a blurred picture, in which none of the points or lines of the one would be united with the points or lines of the other; but were we to look at the combination with both eyes the blurred picture would start into relief, the eyes uniting in succession the separate points and lines of which it is composed.

Now, since, in the stereoscope, when looked into with two eyes, we see the picture in relief with the same accu

racy as, in ordinary binocular vision, we see the same object in relief by uniting on the retina two pictures exactly the same as the binocular ones, the mere statement of this fact has been regarded as the theory of the stereoscope. We shall see, however, that it is not, and that it remains to be explained, more minutely than we have done in Chapter III., both how we see objects in relief in ordinary binocular vision, and how we see them in the same relief by uniting ocularly, or in the stereoscope, two dissimilar images of them.

Before proceeding, however, to this subject, we must explain the manner in which half and quarter lenses unite the two dissimilar pictures.

In Fig. 17 is shewn a semi-lens MN, with its section

M'N. If we look at any object successively through the portions A A'A" in the semi-lens MN, corresponding to aa'a" in the section M'N', which is the same as in a quarter-lens, the object will be magnified equally in all of them, but it will be more displaced, or more refracted, towards N, by looking through A' or a' than through A or a, and most of all by looking through A" or a", the refraction being greatest at A" or a", less at A' or a', and still less at A or a. By means of a semi-lens, or a quarter of a lens of the size of MN, we can,

with an aperture of the size of A, obtain three different degrees of displacement or refraction, without any change of the magnifying power.

If we use a thicker lens, as shewn at M'N'nm, keeping the curvature of the surface the same, we increase the refracting angle at its margin N'n, we can produce any degree of displacement we require, either for the purposes of experiment, or for the duplication of large binocular pictures.

When two half or quarter lenses are used as a stereoscope, the displacement of the two pictures is produced in the manner shewn in Fig. 18, where LL is the lens for the left eye E, and L'L' that for the right eye E', placed so that the middle points, no, n'o', of each are 2 inches distant, like the two eyes. The two binocular pictures which are to be united are shewn at ab, AB, and placed at nearly the same distance. The pictures being fixed in the focus of the lenses, the pencils ano, A'n'o', bno, B'n'o', will be refracted at the points n, o, n',o', and at their points of incidence on the second surface, so as to enter the eyes, E, E', in parallel directions, though not shewn in the Figure. The points a, A, of one of the pictures, will therefore be seen distinctly in the direction of the refracted ray—that is, the pencils an, ao, issuing from a', will be seen as if they came from a', and the pencils bn, bo, as if they came from b', so that ab will be transferred by refraction to a'b'. In like manner, the picture A B will be transferred by refraction to A'B', and thus united with a'b'.

The pictures ab, AB thus united are merely circles, and will therefore be seen as a single circle at A'B'. But if we suppose ab to be the base of the frustum of a cone, and cd its summit, as seen by the left eye, and the circles AB, CD

to represent the base and summit of the same solid as seen by the right eye, then it is obvious that when the pictures of

cd and CD are similarly displaced or refracted by the lenses LL L'L', so that cc' is equal to aA' and DD' to BB', the circles will not be united, but will overlap one another as at