26 STEREOTROPISM OE THE DOGFISH 



and I found that the amount of bending necessary to produce any 

 reflex movement in this way was vastly more than could be caused 

 by the contact stimuli. 



3. Retinal Stimuli. — I have already mentioned the fact that dogfish 

 swimming about in the concrete tank often keep close to the wall, 

 with the dorsal fins slightly flexed and the body tilted so that the 

 belly is turned slightly outward toward the wall. It was possible 

 that this position was induced by the moving image of the wall upon 

 the retina. In making the strokes used as contact stimuli I naturally 

 passed my hand close to the eye of the stimulated side. It was possi- 

 ble that the flexion of the dorsal fins might be due in reality to the 

 image (or the shadow) of the hand upon the retina. I found in fact 

 that when the fish was placed on the board parallel to the window, so 

 that the left eye was toward the light, passing the hand between the 

 window and the animal's left eye often caused a definite flexing of the 

 first dorsal fin to the left, but I could never by this means get a move- 

 ment of the other dorsal or of the paired fins. In order, however, to 

 test this matter farther I made many experiments in which the eyes 

 were covered with thick pads of wet absorbent cotton and found no 

 apparent change in the responses to contact stimuli. 



CONCLUSIONS. 



In the majority of instances the regularity of the responses to the 

 finger strokes and to the forceps scratches is no more remarkable than 

 the definiteness of the change from the one kind of response to the 

 other. The two kinds of stimuli mentioned differed sufficiently to 

 give reactions of opposite sign. An occasional animal, however, 

 reacted very feebly or not at all to the finger stroke, and gave "posi- 

 tive" reactions to fairly strong forceps scratches, in fact in a few 

 instances no negative reaction was obtained. Other specimens gave 

 positive reactions to moderately strong forceps scratches (stimuli 

 which in the great majority of specimens would cause a lively nega- 

 tive reaction), but gave the negative reaction when still more force 

 was applied. On the other hand a few gave only the negative reac- 

 tion to any eft'ective stimulus. 



It was noticeable, too, that what constituted a "weak" or a 

 " strong " stimulus depended upon the physiological state of the individ- 



