THREE years ago I published some new methods of stereoscopy and pointed out that from a mathematical point of view Roentgen-ray stereoscopy is a much simpler problem than ordinary stereoscopy with a photographic camera with lenses. With Roentgen rays the object and its projection are on the same side of the centre of projection, whereas with the camera they lie one on either side of this point, and the image is therefore reversed. With the camera, owing to the position of the conjugate loci,

more »

... ratio of the distances is fixed, but with the Roentgen rays this is not the case. With the rays, the plate may be at any angle, althoi~gh I only considered the " normal " case, so as to simplify the construction. I pointed out the mistake of applying the laws of lens stereoscopy to Roentgen stereoscopy, since, in my opinion, the simplest way is to derive ordinary stereoscopy from Roentgen stereoscopy. The so-called " pin-hole camera" is a transition stage between these two; in this the law of conjugate loci does not apply, and there is no principal axis.