RHEOTAXIS IN ISOPODA 265 



3 decided movement of the anterior end of the body or 



crawhng one or two steps 



4 more extended crawling covering about a centimeter 



5 crawhng over a centimeter 



6 a definite jump follow^ed by rapid running 



The use of these numbers in recording the reactions is illustrated 

 in table 3. 



Banta records a mixed type of response. In number four of 

 his classes of reaction he states that the isopod moved antennae 

 or other appendages or bent the anterior end of the body and 

 followed this by crawling. Reactions of this kind were not given 

 under the conditions of my experiment and may have been due 

 to the repeated rebounds acting as a cumulativ^e stimulus. Con- 

 cussion occurring while the animal was crawling usually caused 

 a cessation of action. Often when the isopod was quiet the 

 only reaction given was a settling of the body. 



Table 3 shows some tests for one isopod made on successive 

 days. The test recorded in column 7 was made immediately 

 after that in column 6 but with a ball weighing only two grams 

 and hence it developed a momentum of only 626 C. G. S. units. 

 This appears to be near the threshold of stimulation when the 

 isopod is as free as possible from thigmotactic interference. 



The tests recorded in columns 5 and 9 immediately followed 

 those given in columns 4 and 8 respectively. The difference was 

 that in the first series of trials the isopods were resting more or 

 less closely in the angle at the edge of the dish, while in the 

 second they had been stimulated until they came to rest free 

 from the edge. In column 5 there is some evidence that part 

 of the increased response was due to this recent mechanical 

 stimulation. By comparing these reactions it becomes apparent 

 that the position of the isopod with reference to the angles of 

 the container may affect their response to a shock stimulus. 

 The position of the isopod in the containers used cannot be 

 readily controlled by the experimenter without increased mechan- 

 ical stimulation. Suitable dishes with hemispherical inner sur- 

 faces were not obtainable when these experiments were per- 

 formed. Since the error from this cause is the same for normal 



