MOVEMENTS OF HERRING AND OTHER MARINE FISHES. 325 



the solutes and other significant data. The fish was observed 

 continuously for twenty or more minutes. 



In order to maintain a constant flow the water was introduced 

 into the tank by means of siphons from cans on the top of the 

 hood with a 74 cm. head. Connected with one of the two cans 

 was an incHned plank trough 420 cm. by 25 cm. for the purpose 

 of aerating water before it entered the can if so desired. 



By the method just described it is possible to obtain unusually 

 accurate data on the factors influencing the movements of 

 fishes. According to Marsh and Cobb ('07) a great difficulty 

 in the herring fishery is the erratic movements of the fish. 

 Schools may visit a bay for three or four years, in succession, and 

 then without any apparent reason, avoid it for a season or two 

 or altogether. Bertham ('97) noted a possible relation between 

 the abundance of these fishes and weather and suggests that 

 climatic cause may have more to do with the failure of some 

 branches of the fisheries than is generally believed. He attri- 

 buted the failure of the fisheries of Cape Benton to the occurrence 

 of severe east and northeast storms during the running season. 

 It is not clear what the effect of such storms may be, but they 

 chiefly affect the dissolved content of the water. Johnstone, '08, 

 page 246, says that it is now nearly certain that the shoaling 

 migrations of the herring of Europe are to be associated with the 

 salinity and temperature of the sea. It is evident from these 

 experiments that acidity and alkalinity are more important than 

 salinity and the solution of the problem will come from a careful 

 study of the reactions of fishes along with a similar study of 

 hydrographic conditions. 



2. Reactions to Temperature. 



These fishes are remarkably sensitive to differences in tem- 

 perature. We obtained good reactions with a difference of 

 0.6° C. in the length of the tank. Fair reactions were obtained 

 with differences of 0.5° C. and since the fishes often turned around 

 near the center it appears that they recognized a difference of 

 0.2° C. In graph i, Chart I., we show the reaction of fish in a 

 gradient of 0.6° C. (compare with graph 2 — control). The fish 

 was taken from sea water at 10.9° and the experiment performed 



