105 



images. For this experiment he placed under each tube of the ste- 

 reoscope a diagram composed of vertical and horizontal lines crossing 

 each other. 



The two diagrams, perfectly identical when seen in the stereoscope, 

 coalesced and formed only one figure ; but although each diagram, 

 when seen separately by its corresponding eye, appeared perfectly 

 flat, still the coalescing image of the two presented a surface con- 

 spicuously concave ; consequently there was no doubt that the same 

 illusion observed in photographic pictures was due only to the effect 

 of the stereoscope. This experiment was decisive, and it remained 

 to discover how the illusion was produced. The investigations 

 showed that the phenomenon, which is a defect detrimental to the 

 beauty and correctness of the stereoscopic representations and un- 

 avoidable in the refracting stereoscope, is a plain illustration of the 

 cause of relief and distance, and yield the clearest explanation of the 

 stereoscopic illusion, proving that it is founded on the true prin- 

 ciples of natural binocular vision. 



When we look through a prism placed near the eye at a straight 

 line, the refracting edge of the prism being parallel with the straight 

 line, that line is refracted laterally and appears bent, with its con- 

 cave side turned to the thin edge of the prism. The two tubes of 

 the stereoscope being supplied with semi-lenses acting as prisms, each 

 lens bends all vertical straight lines, and the concave sides of these 

 lines are turned towards the thin edges of the lenses, and conse- 

 quently towards each other. When we examine in the stereoscope 

 two curved lines having their concave sides turned towards each 

 other, the result of the coalescing of these two lines is a concave line, 

 the extremities appearing nearer and the centre further. If the con- 

 vex sides are turned towards each other, the result of the coalescence 

 is a convex line, the extremities appearing further and the centre 

 nearer. By the same reason, if straight lines are bent by the pris- 

 matic refraction of the two semi-lenses, as the bending is effected so 

 that the concave sides are turned towards each other, the result is 

 by coalescence a concave line. The two photographic images will 

 have all their vertical lines bent in the same manner, and the stereo- 

 scope will give the illusion of a picture represented on a concave 

 surface. 



When we look at natural objects, the optical axes have to con- 



