180 S. O. MAST 



on the longitudinal axis (fig. 10), and this results in alternate 

 decrease and increase in the illumination of the optic nerve-endings ; 

 when the inner surface of the cup faces the light the nerve-endings 

 are fully exposed and when the outer surface faces it, they are 

 shaded by the wall of the cup which is very nearly opaque. 

 Whenever the illumination of the optic nerve-endings decreases 

 the tail bends and the anterior end turns sharply toward the 

 ocular side, and whenever it increases the tail bends in the oppo- 

 site direction (fig. 10, ^). The eye is so situated that when the 

 body rotates from a position in which the eye is on the shaded 

 side to one in which it is on the illuminated side or a little be- 

 yond (fig. 10, 3) the inner surface of the eye-cup and the optic 

 nerve-endings become shaded. The tail then immediately 

 bends sharply toward the light and downward, but owing to the 

 extremely rapid rotation of the body the anterior end is turned 

 from the light (fig. 10, 5). But as the body rotates the eye is 

 again carried to the shaded side, and as soon as it passes a 

 little beyond, the inner surface of the cup becomes exposed to 

 the light, the illumination of the optic nerve-endings increases 

 and the tail responds by bending toward the abocular side, i.e., 

 toward the light and somewhat downward. Owing to the rapid 

 rotation, however, the anterior end is again turned from the 

 light just as in the preceding half of the rotation (fig. 10, 7). 

 Thus it continues until it is directed from the light and rotation 

 on the longitudinal axis no longer produces changes in the illumi- 

 nation of the optic nerve-endings that exceed the threshold. 

 In this direction it continues, not because it is continuously 

 stimulated and held upon its course like a weather-vane in the 

 wind, or like Hammond's 'mechanical dog,' as maintained by 

 some (Loeb, '20, p. 68,) but because it tends to continue on a 

 direct path in the absence of directive stimulation. And if, 

 owing to obstruction or internal stimulation, it is thrown out of 

 this course so that rotation again results in changes in illumi- 

 nation of the optic nerve-ending, the reactions just de- 

 scribed are repeated and the organism again becomes oriented. 

 In ]3ositive specimens, as we have seen, the tail, in response 

 to changes in the illumination of the optic nerve-endings, bends 



