If we place a small red and violet disc, like the smallest wafer, beside one another, so that the line joining their centres is perpendicular to the line joining the eyes, and suppose that rays from both enter the eyes with their optical axes parallel, it is obvious that the distance between the violet images on each retina will be less than the distance between the red images, and consequently the eyes will require to converge their axes to a nearer point in order to unite the red images, than in order to unite the violet images. The red images will therefore appear at this nearer point of convergence, just as, in the lenticular stereoscope, the more distant pair of points in the dissimilar images appear when united nearer to the eye. By the two eyes alone, therefore, we obtain a certain, though a small degree of relief from colours. With the lens LL, however, the effect is greatly increased, and we have the sum of the two effects.

It is,

From these observations, it is manifest that the reverse effect must be produced by a concave lens, or by the common stereoscope, when two coloured objects are employed or united. The blue part of the object will be seen nearer the observer, and the red part of it more remote. however, a curious fact, and one which appeared difficult to explain, that in the stereoscope the colour-relief was not brought out as might have been expected. Sometimes the red was nearest the eye, and sometimes the blue, and sometimes the object appeared without any relief. The cause of this is, that the colour-relief given by the common stereoscope was the opposite of that given by the eye, and it was only the difference of these effects that ought to have been observed; and though the influence of the eyes was an

inferior one, it often acted alone, and sometimes ceased to act at all, in virtue of that property of vision by which we see only with one eye when we are looking with two.

In the chromatic stereoscope, Fig. 42, the intermediate part mn of the lens is of no use, so that out of the margin of a lens upwards of 21 inches in diameter, we may cut a dozen of portions capable of making as many instruments. These portions, however, a little larger only than the pupil of the eye, must be placed in the same position as in Fig. 42.

All the effects which we have described are greatly increased by using lenses of highly-dispersing flint glass, oil of cassia, and other fluids of a great dispersive power, and avoiding the use of compound colours in the objects placed in the stereoscope.

It is an obvious result of these observations, that in painting, and in coloured decorations of all kinds, the red or less refrangible colours should be given to the prominent parts of the object to be represented, and the blue or more refrangible colours to the background and the parts of the objects that are to retire from the eye.

11. The Microscope Stereoscope.

The lenticular form of the stereoscope is admirably fitted for its application to small and microscopic objects. The first instruments of this kind were constructed by myself with quarter-inch lenses, and were 3 inches long and only 1 and 1 deep.1 They may be carried in the pocket, and exhibit all the properties of the instrument to the greatest advantage. The mode of constructing and using the instru ment is precisely the same as in the common stereoscope;

1 Phil. Mag., Jan. 1852, vol. iii. p. 19.

I

but in taking the dissimilar pictures, we must use either a small binocular camera, which will give considerably magnified representations of the objects, or we must procure them from the compound microscope. The pictures may be obtained with a small single camera, by first taking one picture, and then shifting the object in the focus of the lens, through a space corresponding with the binocular angle. To find this space, which we may call x, make d the distance of the object from the lens, n the number of times it is to be magnified, or the distance of the image behind the lens, and D the distance of the eyes; then we shall have nd: d=D: x, and x =

Ꭰ

that is, the space is equal to the distance between the eyes divided by the magnifying power.

With the binocular microscope of Professor Riddell,1 and the same instrument as improved by M. Nachet, binocular pictures are obtained directly by having them drawn, as Professor Riddell suggests, by the camera lucida, but it would be preferable to take them photographically.

Portraits for lockets or rings might be put into a very small stereoscope, by folding the one lens back upon the other.

1 American Journal of Science, 1852, vol. xv. p. 68.

CHAPTER VIII.

METHOD OF TAKING PICTURES FOR THE STEREOSCOPE.

HOWEVER perfect be the stereoscope which we employ, the effect which it produces depends upon the accuracy with which the binocular pictures are prepared. The pictures required for the stereoscope may be arranged in four classes:

1. The representations of geometrical solids as seen with two eyes.

2. Portraits, or groups of portraits, taken from living persons or animals.

3. Landscapes, buildings, and machines or instruments. 4. Solids of all kinds, the productions of nature or of art.

Geometrical Solids.

Representations of geometrical solids, were, as we have already seen, the only objects which for many years were employed in the reflecting stereoscope. The figures thus used are so well known that it is unnecessary to devote much space to their consideration. For ordinary purposes they may be drawn by the hand, and composed of squares, rectangles, and circles, representing quadrangular pyramids, truncated, or terminating in a point, cones, pyramids with polygonal bases, or more complex forms in which raised

pyramids or cones rise out of quadrangular or conical hollows. All these figures may be drawn by the hand, and will produce solid forms sufficiently striking to illustrate the properties of the stereoscope, though not accurate representations of any actual solid seen by binocular vision.

If one of the binocular pictures is not equal to the other in its base or summit, and if the lines of the one are made crooked, it is curious to observe how the appearance of the resulting solid is still maintained and varied.

The following method of drawing upon a plane the dissimilar representations of solids, will give results in the stereoscope that are perfectly correct :

Let L, R, Fig. 43, be the left and right eye, and a the middle point between them. Let MN be the plane on