346 PROCEEDINGS OF THE AMERICAN ACADEMY. 



were also tested and compared. The differences in the spectral com- 

 ponents V7ere found to be so slight as to be negligible. A sheet of clear 

 glass, equal in thickness to the ground glass of the large light, was at 

 first placed in front of the small light, in order to make the conditions 

 of the two as similar as possible ; but it was found to be of little use, 

 and was of such obvious disadvantage, because of the considerable dis- 

 persal and reflection of the light which it caused, that it was later dis- 

 pensed with entirely, and consequently is not represented in the figures 

 . of the apparatus. 



The vertical light (F), which has been described, was used only in 

 connection with the small light (Sm) in studying chiefly the reactions 

 of animals to lights of equal intensity, but coming from different direc- 

 tions, — in this case one light striking the animal from directly above, 

 the other coming to it horizontally. This combination of lights was 

 also employed in studying the reactions to shadows and to sudden 

 differences of intensity. These experiments will be described in another 

 paper. 



Other appliances and conveniences will be mentioned in their proper 

 places. It should be stated here, however, that screens were provided 

 by means of which the light from any of the lamps could be immedi- 

 ately shut off from the working position at will. 



At the point on the table exactly midway between the lights was 

 painted a white line, and this may be spoken of as the 7iormal axis of 

 the apparatus {a', Figure 2), since it is at right angles to the directive 

 axis already mentioned. At the beginning of each trial in an experi- 

 ment the subject was usually placed, headed in one direction or the 

 other, on this line, its long axis coinciding with the line. For the sake 

 of brevity in description, this is called the normal position. Any line 

 running parallel to the normal axis — nearer to one light or the other — 

 may be said to run in a normal direction. The centre of the ivorking 

 position was at the point where the normal and directive axes crossed, 

 and the working area included a small horizontal surface, the extent of 

 which was determined by the distance the animals were allowed to 

 move. Three concentric circles, the inner with a radius of 5 cm., the 

 second with a radius of 10 cm., and the outer with a radius of 15 cm., 

 were described with their centres at this point. Each of the circles was 

 divided by short cross lines into arcs of 10°, and by means of this 

 device the angles at which the animals deviated from the normal axis 

 and crossed any of the circles could be read directly with ease and 

 accuracy. 



