196 LIGBT AND THE BEHAVIOR OF ORGANISMS 



is true with regard to the fly larvae. In Stentor every rota- 

 tion on the spiral course may be considered a trial move- 

 ment. If the organism is not oriented it swerves a little 

 farther from the source of light in each rotation after the 

 oral side is turned toward the light, until this side is equally 

 exposed throughout the entire rotation. Just so every lat- 

 eral movement of the fly larvae may be considered a trial 

 movement. If the organism is not oriented the anterior 

 end becomes much more fully exposed when it is turned 

 toward the light than when it faces in the opposite direc- 

 tion. This produces a stimulation and causes it to be 

 turned farther than usual in the opposite direction, but it 

 is swung back again, receives another stimulation, and is 

 turned still farther from the light. Thus the organism may 

 be considered to try different positions by swinging the 

 anterior end back and forth. This trial process does not 

 cease after the organism is oriented; the anterior end con- 

 tinues to swing from side to side. If it is subjected to but 

 little difference of light intensity as it swings from side to 

 side, there is no response and the organism continues as it 

 is directed, but if it is subjected to considerable difference 

 of intensity it responds and turns as described above. 



The fly larva presents an excellent example of an organ- 

 ism guided fairly directly on its course by successive trial 

 movements, and shows again that the mere fact of accurate 

 orientation is not a satisfactory criterion of direct orienta- 

 tion. Of course it is not necessary to assume that this 

 organism consciously tries different positions in the process 

 of orientation. 



In how far do the reactions of the fly larvae agree with 

 the explanation Holmes presented with reference to them ? 

 Is "orientation in these forms . . . essentially a selection of 

 favorable chance variations of action and following them 

 up" ? The answer to this question depends entirely upon 

 what is meant by chance variations. It is therefore evi- 

 dent that a statement with reference to it would add little 

 or nothing to our analysis. 



