SUGGESTED IMPROVEMENT IN BINOCULAR MICROSCOPES. 393 



impinge on the object, and form the image that is brought to 

 the right and left eye respectively. Why stereoscopic effect fails 

 with high-power binoculars is a subject of some complexity, and, 

 indeed, does not appear to have been thoroughly investigated as 

 yet. We know that it is influenced by the fact that we cannot 

 fill our objectives with light in the way we are able to do with 

 lower powers ; also, that the small depth of the object it is possible 

 to view at any one time limits its scope. The chief reasons, 

 however, I believe, will be found to be in, — 



1. Exaggerated parallactic displacements as we proceed to 

 higher power. They follow from the same causes as those which 

 produce the increasing super-magnification of the depth dimension. 



2. The peculiar way in which the influence of diffraction 

 makes itself felt when we are dealing with separated portions 

 of objectives of large N.A., and have to consider out-of-focus 

 images obtained from comparatively narrow cones of light. 

 But further inquiry is necessary before it is possible to adequately 

 enter into these matters. 



Before concluding this paper, just a few words on the subject 

 of stereoscopic photo-micrography. As already mentioned, 

 Mr. Taverner's experimental results in this branch of microscopic 

 work led to the present inquiry and adaptation of the method 

 to binoculars. The conditions in the main are similar, but we 

 have the following points of difference : — 



1. The penetrating power in the case of photography depends 

 wholly on the size and configuration of the diffusion discs, as the 

 eye accommodation is absent. The improvement arising from 

 substitution of circular stops for the semicircular half -objective 

 aperture is even, therefore, slightly more marked. 



2. Since we are able to use the parts of the objective wanted 

 for the right and left eye pictures successively, instead of 

 simultaneously, we have greater scope for adjusting the paral- 

 lactic displacements, so as to obtain the correct impression of 

 the perspective. We may, for example, use overlapping portions 

 of an objective, as in Fig. 9. In the latter the amount of 

 parallax (represented by the distance between the centres of 

 the circular stops) is decreased, and yet good resolving power 

 of the plane in true focus need not be sacrificed, as the size of 

 the opening can be left sufficiently large to yield good resolution 

 of the object. 



