66 Bulleti7i of Lahoratoties of Denison University. [Voi.xii 



fresh lot of Ameiurus fry under observation and have already 

 verified many of the conclusions reached with the first lot. But 

 this second collection of fishes has not at the time when this 

 report is submitted been in captivity long enough to become 

 sufficiently accustomed to their new surroundings to feed freely 

 and fearlessly. After some months of further preliminary ob- 

 servation I hope to carry on experiments which may shed some 

 light on the sense of smell in these fishes. But this must be 

 reserved for a later report. 



We must content ourselves at the present time, then, with 

 the inference that the sense of smell plays at least a small part 

 in these reactions, for the animals became slightly restless in 

 the proximity of the stimulus, though they were not in contact 

 with it ; this, however, appears never to provoke a definite re- 

 action of seizing the food, but merely a vague reaction in search 

 of food. On the other hand, physical contact with the irrita- 

 ting substance causes a definite and precise reaction which is 

 practically constant. This points either to touch or to taste. 



To test the relative part played by stimulation of these 

 two sets of sense organs the following series of experiments 

 was performed. A half a dozen fish in an aquarium were tested 

 a score of times with fresh meat on the tip of a wire, as in pre- 

 vious cases. The reaction was obtained uniformly, no matter 

 what part of the body or head was touched. Half an hour 

 after the close of these experiments a bit of cotton wool was 

 wound around the tip of a wire and the fishes were tested with 

 this exactly as they had been with the meat. For the first six 

 trials the barblets only were touched. The fish in each case 

 turned and seized the cotton as promptly as the meat had been 

 taken. The cotton would be immediately dropped. After a 

 few more trials the fishes would generally turn when touched, 

 but would check their movement before the cotton was actually 

 taken into the mouth. Several specimens were now tested on 

 the trunk with the cotton. One or two turned completely 

 around and took the cotton, but generally there was a slight 

 movement only toward the cotton, which was checked before 

 the cotton was reached. After a few further tests, the fishes 



