procession of Cnethocampa pinivora. 579 



towards the ground. In spite of much search I never found one 

 reaching to the ground and very rarely within a yard from it. 

 There is little difficulty in seeing the thread, which glistens as it 

 lies adhering to the flakes of bark. Moreover it is very durable 

 and does not seem to suffer much from wind or rain, as shown by 

 the matting over of the upper branches and the durability of the 

 nest itself, which is essentially a winter dwelling. Again, the 

 thread left by a procession is the aggregate secretion of all the 

 larvae ; Fabre calls attention to its thickness when formed by 

 a long procession, and Edwards records that 150 larvae may make 

 a thread nearly 3 mm. wide. Such a thread is very strong and 

 I have lifted the thread of a procession clear of the ground for two 

 feet without breaking it, though it was loaded with adhering sand 

 particles. Now the threads which I found running down the 

 trunks of saplings were thin ones resembling those secreted by 

 single larvae, and certainly did not suggest processions frequently 

 leaving and returning to the tree. There seemed no ground for 

 thinking that threads near the base of the trunk are especially 

 likely to be swept off accidentally, by the movements of under- 

 growth in wind for instance. It must be admitted that the 

 examination was on trees bearing in most cases deserted nests 

 and perhaps some unknown cause had removed the threads near 

 the ground since the final procession had left the tree, but if a 

 thread near the base lasts as long as those on the branches, that 

 of the final procession should have been found. But on the whole 

 the impression left by the threads on the bark is that the larvae 

 do not habitually leave the tree and return to it. 



If the thread is really a guide to return to the nest tree the 

 leader of a homeward bound procession must make use of it 

 by coming across its own thread and recognising it as the one 

 to follow. Edwards sought for information on this point by 

 placing a frayed-out piece of artificial silk or the natural thread 

 of another procession in the leader's path — this was disregarded 

 entirely. I repeated Fabre's experiment of picking up the thread 

 of a procession with forceps and looping it round so as to be 

 encountered by the leader. On each occasion, and the experi- 

 ment was made with different processions, the leader crossed the 

 thread without regarding it more than any other small obstacle 

 he encountered. This accords with Fabre's experience (loc. cit. 

 p. 339), when endeavouring to make the larvae march in a closed 

 circle. If a larva can distinguish its own thread among the 

 numerous ones which are seen on fine days crossing the sand 

 in all directions, it appears probable that it would show some 

 recognition of it so soon after its formation as this, even^ if its 

 psychological condition were not in favour of a return to the nest 

 tree at the moment. It can only be said that much more 



VOL. xy. PT. VI. 38 



