40 



Here we have positive proof that the distinguishing habit of this 

 larva can be performed without violating the principle of bifurcator 

 control. Until some more simple explanation of the fact that the 

 arms are thus definitely correlated to the length of the first nine seg- 

 ments can be brought forward, we may, I think, accept as true in 

 those larvae possessing functionally active bifurcators the following 

 hypotheses : (i) The bifurcating hairs are sensory organs. (2) The 

 feeling-tone set up in the brain of the insect when they are stimulated 

 is pleasurable. (3) In consequence of this feeling-tone the larva avoids 

 doing anything that would put an end to this sensation, and so never 

 exposes more than the first nine segments. In other words the bifurcators 

 generate a new instinct. 



Effects of bifurca The bifurcator instinct as we have seen controls the movements of 



the larva in and out of its case. It regulates the length of each arm 

 of the net, and consequently the diameter of the net as a whole. This 

 is a very important point, for if it was not for the bifurcator instinct the 

 net might be 3 or 4 mms. wider than it actually is and so be able to 

 bring to the larva a larger supply of food. But if the larva not having 

 this instinct, came right out of its case as far as possible, and so built 

 full length arms, its whole body would be exposed to the full force of 

 the current ; but by slightly restricting the size of the net comparative 

 safety is ensured to the animal. 



The remarks apply to both Chiron owns pusio and the stalk-cased 

 larva, but in the latter not only is the length of the arm controlled by 

 the bifurcator instinct, but the length of the case is also affected in the 

 same way. For this larva stretches out over the edge of its case and 

 reaching backwards tears away material from the posterior end to add 

 to the anterior end. The bifurcator instinct will not allow it to expose 

 the last four segments, consequently it cannot reach very far down the 

 outside of the case to begin its destructive and constructive work, so 

 that the actual case existing at any moment cannot be longer than the 

 head and the first eight segments, if we assume the ninth segment to 

 occupy the bend and the last four to be within the tube. 



