326 VICTOR E. SHELFORD AND EDWIN B. POWERS. 



at 12.8° and 13.2°. It will be noted that the fish showed a prefer- 

 ence for the higher temperature. Eleven experiments were 

 performed with herring and in seven cases the fishes showed a 

 preference for the warmer water and in three cases for the colder. 

 One did not show any marked preference. The differences were 

 too slight to be of great significance in determining whether the 

 fishes move into warmer or colder water but show a great sensi- 

 tiveness. Thus temperature may play an important role in the 

 movements of fishes. 



It will be noted by reference to the graph, that the fish moved 

 into the colder water several times as if trying out the entire 

 tank and then turned back periodically from the colder end. 

 In the control where there was no flow or difference in tempera- 

 ture the fish turned back from both ends at times but by chance 

 as shown by other controls, turned a little more often from the 

 end corresponding to the cold end of the experiment due perhaps 

 to difference. 



3. Hydrogen Sulfide. 



The animals turned back sharply from all concentrations not 

 great enough to cause intoxication as shown in graph 3, Chart I. 

 (compare with control graph 4) . In this experiment the hydrogen 

 sulfide was only 4.5 c.c. per 1. and the fishes avoided it sharply 

 and after trying out the tank turned about at a point where the 

 concentration could not be more than one tenth of that at the 

 treated water end or about equal to that under the Ulva on the 

 south side of Brown's Island (p. 319). 



This experiment is typical of several and the fishes are thus 

 seen to be able to orient with reference to an increase in the 

 solute and to turn back from it very sharply. 



The control (graph 4) to this experiment is symmetrical, 

 there being turning from both ends in equal number. It shows 

 the reaction of the fishes when no stimuli are encountered in the 

 tank. 



4. Reactions to Salinity, Acidity and Alkalinity. 



As noted above, the fresh water of the laboratory was from 

 deep wells and not good for biological work. It was alkaline, 

 containing no free carbon dioxide, 24.2 c.c. per 1. half bound and 



