24 VICTOR E. SHELFORD AND W. C. ALLEE 



containing 150 cc. or more of carbon dioxide and only lcc. of 

 oxygen per liter, very striking evidences of stimulation appear 

 almost instantly. In 47 cc. of carbon dioxide per liter in boiled 

 water nearly all the fishes showed evidence of stimulation at once. 

 They would start gulping before reaching the end of the tank 

 on the first entrance of the solute. In only a few of the cases 

 was this delayed as long as 50 seconds. In from 20 to 60 cc. of 

 carbon dioxide per liter in tap water the fishes showed similar 

 stimulation in from two to ten seconds after entering the high 

 concentration. In boiled water some fishes showed increased 

 activity, gulping within ten seconds, but such manifestations 

 were frequently delayed for nearly a minute and were quite 

 variable in intensity. In acetic acid the evidences of stimu- 

 lation were similar to those in carbon dioxide and in ammonia 

 some of them were noticeable. 



The fishes undoubtedly sense the solutes upon entering them. 

 For this, they give evidence by the following activities: A definite 

 reflex was often given by Abramis, Notropis, Hypopsis and 

 Lepomis the first time they entered the modified water. The 

 fish suddenly stopped, backed quickly a few millimeters and then 

 started ahead again, often repeating the reflex before going 

 farther forward. In the earlier paper, we called this the backing- 

 starting reaction. This may be due to stimulation of the nostrils. 

 Sheldon ('09, p. 278) states that stimulation of the nostrils of 

 the dog fish resulted in a quick jerk of the head. There was ac- 

 celeration or increased vigor of movement of fins, tail or body 

 which began at once or after a very short time. Sheldon found 

 that the application of solutions to these parts caused them to 

 be moved. The opercles were lifted, the lower jaw protruded, 

 or the mouth moved in a manner characterized as coughing, 

 gulping or yawning. Sheldon found that stimulation of the mouth 

 or spiracle gave rise to violent gulps. In our experiments these 

 reactions occurred singly or in combination. The time necessary 

 to produce them was variable, but depended upon the strength 

 of the stimulus, which confirms further observations by Sheldon 

 ('09). 



As further evidence of the quick sensing of the stimuli, Hybopsis 

 turned back the first time the gradient was tried in nine cases, 

 Notropis in three cases, Ameiurus in four cases, Umbra in one 

 case, and Abramis in one case. With a single exception carbon 



