126 THE WONDERS OF INSTINCT 



order would still be perfect: the second touches and fol- 

 lows the first. 



By February I have processions of all lengths in the 

 greenhouse. What tricks can I play upon them? I see 

 only two: to do away with the leader; and to cut the 

 thread. 



The suppression of the leader of the file produces 

 nothing striking. If the thing is done without creating 

 a disturbance, the procession does not alter its ways at 

 all. The second caterpillar, promoted to captain, knows 

 the duties of his rank of¥-hand: he selects and leads, or 

 rather he hesitates and gropes. 



The breaking of the silk ribbon is not very important 

 either. I remove a caterpillar from the middle of the 

 file. With my scissors, so as not to cause a commotion 

 in the ranks, I cut the piece of ribbon on which he stood 

 and clear away every thread of it. As a result of this 

 breach, the procession acquires two marching leaders, 

 each independent of the other. It may be that the one 

 in the rear joins the file ahead of him, from which he 

 is separated by but a slender interval ; in that case, things 

 return to their original condition. More frequently, the 

 two parts do not become reunited. In that case, we have 

 two distinct processions, each of which wanders where it 

 pleases and diverges from the other. Nevertheless, both 

 will be able to return to the nest by discovering sooner 

 or later, in the course of their peregrinations, the ribbon 

 on the other side of the break. 



These two experiments are only moderately interest- 

 ing. I have thought out another, one more fertile in 



