32 Dr. ^yndall on tlte Progress of tlie Physical Sciences : 



each eye, so that the hypothenuse surfaces are vertical — best so 

 that the hypothenuse surfaces face each other — and by inclining 

 the prisms, the images of the two drawings placed side by side 

 are brought to coincidence. This stereoscope, like the eye, is 

 applicable to all drawings ; when the prisms are fixed in cylin- 

 drical tubes and attached to a stand, the arrangement is exceed- 

 ingly convenient. , .> 



e. Mirror-stereoscope mth two Drawings and a plane metallic 

 mirror or reflecting prism. 



The drawing intended for the left eye is laid horizontal, and 

 is viewed with the naked left eye. Before the right eye is held 

 a small metallic mirror or a reflecting prism ; and in the case of 

 simple figures, another copy of the same drawing is observed in 

 a vertical plane. For an unsymmetrical figm*e, a drawing which 

 is the reflected image of that intended for the right eye must be 

 used. 



f. Double-seeing, as Stereoscope. 



Those who have accustomed themselves to double-seeing, when 

 two stereoscopic drawings are laid side by side, can obtain four 

 images in a direction parallel to a line joining both eyes. The 

 two central images can be brought to coincidence, the relief 

 being thus obtained between its two projections. This experi- 

 ment is, however, so weaiying to the eyes, that its frequent 

 repetition is not to be recommended. The author has made the 

 experiment purely on account of its physiological interest. A most 

 peculiar impression is caused by the union of the two images in 

 the relief. When brought near, they seem to rush together with 

 an accelerated velocity as if they more strongly attracted each 

 other. In a similar manner, when two persons lay their fore- 

 heads together and look into each other's eyes, each observes the 

 eyes of the other to run together into one large eye in the centr^. 

 of the forehead. ^ 



g. Whi/ does the depth of the concave Relief appear greater 

 than t/ie height of the Convex ? 



In the application of the stereoscopes above described, it is 

 observed in a sti-iking manner, particularly when the di'awings 

 are viewed from some distance, that by exchanging the projec- 

 tions the height of the convex relief appears less than the depth 

 of the concave. The plane of the paper on which, as ground 

 surface, the drawing is executed, is seen at the same distance in 

 both cases. That this is the cause of the phsenomenon is proved 

 by the fact, that the side surfaces of the frustum of a pyramid, 

 stereoscopic obseiTcd, seem to be less steeply inclined towards the 

 base when the top of the frustum is nearest the eye, than when 

 it falls at the other side of the base, and is viewed through the 

 hollow prism. The top surface of the frustum is seen in both 



