288 W. C. ALLEE 



efficiency was 2.1. Two hours later the response was 90 per 

 cent positive, 10 per cent no reaction, efficiency 2.8. This last 

 trial was characterized by steady slow movements and quick 

 orienting, the latter seldom taking over five seconds. 



In this case the molt of the posterior end occurred two days 

 later, after 9 p.m. At nine the response was 40 per cent positive, 

 40 per cent negative, 20 per cent indefinite, efficiency 2.6. The 

 next morning the response was 80 per cent positive, 20 per cent 

 negative, efficiency 2.7. On the morning after the next molt, 

 thirty-one days later, this isopod would start positive and then 

 quickly turn negative as if the current pressure against the more 

 sensitive covering were painful. 



Isopods stand higher from the bottom when nearing molting, 

 which is probably due to the increasing stijffness of their legs. 

 This makes resistance to the current more difficult. At this 

 time the posterior legs appear harder to move and may become 

 tangled, thus throwing the isopod as it tries to crawl. Imme- 

 diately after molting they are thrown from their feet more easily 

 than during other parts of the cycle. 



The more gradual molts mentioned above may also affect the 

 rheotactic reaction. Thus in one instance an isopod (no. IV) 

 n^olted the fourth thoracic segment and legs while its response 

 was being tested. This occurred about thirty to thirty-six hours 

 after the anterior molt and was accompanied by a cut of 20 per 

 cent in the positive reaction. Another tendency is indicated by 

 the molt in isopod no. V (table 9). This animal had been in- 

 jured in handling and before molting gave a high percentage of 

 indefinite responses bi^t afterward the percentage of positive reac- 

 tions was suddenly and markedly increased. 



As is seen in the account for the molting observed in isopod 

 no. XI, the effect of the molting lasted for about five hours after 

 the actual ecydsis took place. If the period extended as long 

 beforehand it would make the time during which rheotaxis was 

 affected by the molting process extend over a period of about ten 

 hours. It is conceivable that the general morning tests such as 

 were usually made might not hit this period at all. Sometimes 

 this is the case, as is illustrated in the molting of no. VII (table 



