Prof. Helniholtz on the Telestereoscope. 21 



The essential parts of it are the four mirrors, b, b and c, c, 

 placed vertically in a wooden box, and fixed at an inclination of 

 45° to its longest edge. The exterior mirrors, b, must be large, 



the interior ones, c, may be small, and to avoid distortion all 

 the mirrors must be of the best plate-glass. The light coming 

 from distant objects is reflected twice at a right angle on its way, 

 abed, and falls 9.t ad into the two eyes of the observer. At 

 //diaphragms are placed, to prevent other light entering the 

 eyes than that which has been twice reflected. In the apertures 

 in the box through which the observer looks, it is convenient to 

 place two very feebly concave lenses, of from thirty to forty inches 

 focal distance, because the majority of eyes do not see very 

 ■ distant objects distinctly, which is here the precise point to be 

 attained. Such weak lenses do not hinder normal eyes from 

 seeing distinctly. 



It is in general better to make the mirrors fixed, as is sup- 

 posed to be done in the drawing; for certain physical experi- 

 ments, particularly for the inspection of near objects, it is useful 

 to have the four mirrors turning on vertical axes. 



Each eye of the observer sees in the small mirror the image 

 of the large one at its side, and in the large one the image of 

 the landscape ; the latter, however, he sees in the perspective 

 projections in which they appear from the two large reflectors, 

 by which of course he obtains a much greater difl"erence in the 

 two perspective views than the two eyes of the observer could 

 funiish by looking directly at the landscape. In order to deter- 

 mine accurately the position from which the landscape in this 

 case is regarded, we must seek the images of the observer's eyes 

 given by the two pairs of miirors, which in the figure would lie 

 in the direction ab, beyond b, and at a distance from b = bc + ed. 

 Ky the instrument, therefore, the distance between the observer's 

 eyes is artificially magnified to bb. 



The land^scape appears to the observer like a reduced model. 

 Jt matters not whether concave {rliisscs arc added to the instru- 



