276 W. C. ALLEE 



placed in a bath that would keep the temperature within one 

 degree of that to which they were accustomed. In the field 

 the trials were made in diffuse light, with all other conditions 

 as near those of the laboratory trials as was possible. 



After fifteen minutes a current was produced by stirring with a 

 glass rod about 8 mm. in diameter. In order to secure as even a 

 current as possible the rod was run five times around the pan at a 

 distance of about 4 cm. from the edge and at a uniform rate. The 

 attempt was made to keep from stirring the animals from the 

 bottom. Usually they remained along the edge of the pan and 

 thus were in that part of the current that shows the least spiral 

 tendency. Time was taken with a stop watch for one minute 

 after the stirring stopped and the reaction of each animal for the 

 greater part of the minute was recorded. Thus if an animal went 

 against the current for forty seconds and with it for the remainder 

 of the minute it was counted positive, while if it went with the 

 current half the time and against it the other half it was counted 

 indifferent. At the end of the minute reaction, the current was 

 set up in the reverse direction. The reason for reversing the direc- 

 tion of the current, is that isopods tend to continue in the same 

 direction in which they are started. Thus if by accident nega- 

 tive isopods are all going against the current and at the end of the 

 minute's reaction the current is merely renewed, there would be a 

 tendency to remain positive although in reality their normal reac- 

 tion would be negative or indifferent. If on the other hand the 

 current is reversed, then if they are strongly positive they will 

 reverse their direction thereby showing that they are reacting to 

 the direction of the current and not to chance factors. 



Trials were continued in this fashion until ten consecutive tests 

 had been made. These results were recorded and the percentage 

 of animals going positively, negatively, and indifferently was cal- 

 culated. It is to be regretted that there is as much left to the 

 personal equation as there is in this method, yet it furnishes a 

 fairly stereotyped set of trials that have given closely comparable 

 results. The positive reaction obtained under these conditions, 

 consists of two factors, namely : (1) The percentage of positive re- 

 sponses numerically stated and (2) the positiveness with which 



