METHODS OF STUDYING VISION IN ANIMALS 51 



in a 4" cubical metal box. (At greater expense the Westing- 

 house Nernst lamp, see p. 5, may be mounted on the box 

 in place of the filament. The lamp is far more convenient.) 

 This filament is projected upon the plaster of paris strips by 

 means of a short focus achromatic lens mounted in a tube at- 

 tached to the box. The box can be shifted on a metal arm so 

 as to exclude any one of the beams, and to project in its place 

 a white beam upon the plaster of paris surface. This box is 

 not shown in the diagram. It should be inserted immediately 

 behind the partition D, shown in fig. 7, and consequently in 

 front of the episcotisters. The episcotisters can then be ad- 

 justed so as to decrease the energy of this light to any desired 

 amount. If preferred, a simple vertical optical slit may be mounted 

 in the focus of the projection lens. The intensity of the beam 

 may be conveniently and accurately controlled by the slit with- 

 out the use of the episcotister. 



The details of the mounting of the above optical system will 

 now be taken up with the aid of separate drawings. 



2. Description of the mounting of the apparatus 

 Fig. 7 shows the complete system ready for operation. Since 

 the aim of our report is to present the working parts clearly 

 enough to make duplication possible, the description which 

 follows will be somewhat technical and detailed. Also there 

 will be, necessarily, a certain amount of repetition in the de- 

 scriptions, since fig. 7, showing the whole apparatus, is first 

 taken up, and is followed by the description of the larger draw- 

 ings of certain parts of the system. Figs. 6 and 7 will give a 

 clear idea of the whole apparatus. Figs. 8-12 inclusive are given 

 for the benefit of those who may desire to construct, or at least 

 install the apparatus. Many important dimensions are not 

 given, but the lenses and other optical parts obtainable differ 

 so much that it seemed hardly worth while to give all the dimen- 

 sions. Those which relate specifically to the control of the 

 stimulus are everywhere given. 



As show^n in fig. 7, two rooms, and / are devoted to the 

 apparatus. / is a complete dark-room 25 feet long and 16 feet 

 \\'ide ; O is a smaller room containing a window for the admission 

 of sunlight. When either the arc or Nernst is in use, daylight 

 is excluded from 0. The dividin2: wall between and / is marked 



