METHODS OF STUDYING VISION IN ANIMALS 47 



I . General description of the optical system 

 Fig. 6 is a diagram showing the path of the light. As it is 

 drawn, the arc L is used as a source. The apparatus is adapt- 

 able for use also with sunlight or with a Nernst filament. When 

 sunlight is used, the beam from a heliostat falls directly upon 

 the condenser, K. When the Nernst filament is used it is 

 mounted vertically 12 cm. from the slit S, and is focused upon 

 S by means of a small concavo-convex lens of f aperture and 

 i^" focus (single achromat). Such lenses are ground to order 

 for us by the Wollensak Optical Company, Rochester, N. Y. 



It will be seen as the apparatus is here sketched for use with 

 the arc, that the light from the crater of the positive carbon 

 of the arc L falls upon the face of an achromatic condenser, K. 

 This condenser causes a sharp image of the crater of the positive 

 carbon to fall upon the slit S. The diverging rays of light 

 issuing from this slit are made parallel by the collimating 

 lens C. The parallel cylinder of light next strikes the 

 face of a dense flint glass prism (60°), h refracted, and passing 

 through the objective O, is brought to a focus in a series of 

 colored images of S upon the double slit (at R and G, two beams 

 are shown passing through this slit). 



Immediately behind this double slit tw^o total reflection 

 prisms ni^ and m^ are to be found. As shown in the diagram, 

 they serve the purpose of bringing the red beam near to the 

 green. It is absolutely necessary to keep the two selected beams 

 within about 10 mm. of each other if the reversing and spacing 

 devices which are later described are to work properly. It may 

 be seen that the distance separating the two beams shortly 

 after they issue from the double slit is determined by the sepa- 

 ration existing between m^ and ?«2- Suppose, for example, 

 one beam issues from the red region, and the other from the 

 violet. Now, with the lenses and refracting prism at present 

 in use, the separation would be, without the use of m^ and m^ 

 about 6 cm. In order to bring the red over to the violet and have 

 them in the same relation as are the red and green in the dia- 

 gram, one has to draw prism m2 over toward the violet beam 

 until the red issues from ni^ parallel to the violet and 10 mm. 

 distant from it. This device works equally well for any two 

 selected beams widely separated. On the other hand, if the 



