56 W. C. ALLEE. 



is allowed to dry, will burrow into the moist mud and thus escape 

 death for some time and it is highly probable that they were 

 able to survive the drouth of 1908 in this manner and to take 

 up their usual places though in less than their usual numbers 

 with the late autumn rains. 



Another line of evidence that tends to minimize the importance 

 of rheotactic reaction in the ecology of stream isopods is found 

 in the analysis of the reaction itself. In 6,630 discontinuous 

 minute reaction periods 95 per cent, of which resulted in positive 

 rheotactic responses the average rate of movement was 76 cm. 

 per minute. The current in which these trials were made 

 averaged 472.5 cm. per minute, hence this represents the response 

 to a mild current. When exposed to the continuous action of a 

 straight current of about this strength the most positive stream 

 isopods reversed their reaction and became negative in less than 

 twenty-four hours. As the rate of the current was increased in 

 laboratory tests the isopods came to rest and clung to the bottom, 

 making no effort to advance. This reaction was given before 

 the current became as strong as that found in parts of the County 

 Line Creek. 



These observations show that positive rheotaxis may enable 

 isopods to maintain their position for a time in parts of a stream 

 having a weak current but after continued exposure or in strong 

 currents the clinging reaction becomes the more important one. 

 This is in accord with the fact that isopods are only found in 

 streams having abundant places for support (p. 55). If the 

 clinging reaction is essential for the most positively rheotactic 

 adults, it must be much more so for the early juvenile stages and 

 for adults in the breeding season or during the molting period 

 when the positive rheotactic reaction is either weakened or 

 entirely absent. 



There is no correlation between the seasonal variation in the 

 positiveness of the rheotactic reaction and the seasonal distri- 

 bution, as described for the County Line Creek, but the distri- 

 bution can be fully accounted for by the seasonal variation in 

 places of lodgment and surfaces that give a good foothold for 

 crawling, that is to say in the seasonal changes in the distribution 

 of leaves in the stream. Thus in seasonal as in regional dis- 



