SCIKNTIFIC N0TP:S. 209 



against the weight in favour of the actual habits of eating the ends 

 fi'rst. 



The excursions from the nest for feeding are made at night, but 

 when the nights are cold, say 23° to 41° Fahr., the larva; do not come 

 out, but starvation, if the cold nights continue, has its effect, and the 

 larva' will then come out and feed in daylight, if it be a little cloudy. 



M. Perris' observation referred to seems to show that starvation, 

 when it results from the vast numbers of the larva; having cleared 

 away all the needles of the trees occupied, compels the larva to adven- 

 ture long journeys, in which a frosty night often catches them to their 

 extirpation. 



When after feeding they return to the nest, they do so with great 

 certainty. M. Rabaud leaves the question of how they do so uncertain, 

 he recounts certain observations that demonstrate that they return 

 easily when there is no silken way for them to follow, so that this 

 usually accepted explanation fails. 



He ends the paper with some remarks on the processionary instinct. 

 He says the origin of the instinct is certainly obscure, and even seems 

 to involve a decided danger, he says other social caterpillars disperse 

 over the food plant and return without any processionary procedure, 

 and that individual larvae of T. iiiti/ocdiiipa find the wa}' home quite 

 easily. 



lie appears to study the point entirely with reference to the feeding 

 larva, and therein misses what seems to be the real advantage secured 

 by the processionary instinct. 



This instinct entirely governs the larva at one very important 

 point in its life, that is when it finally leaves the nest and goes to find 

 a place for pupation. It is then that they may be found in procession, 

 by day as well as by night, and when larvrt^ from two different nests 

 readily combine in the same procession. 



That the larvie should keep together in considerable numbers is of 

 the greatest importance, because they enter together some cavity under 

 rubbish, or even underground, and spin their cocoons together in a 

 close mass. 



It is, no doubt, for this event that natural selection has brought 

 the habit to the perfection we observe. The possession of the same 

 instinct in the earlier feeding stage is probably due to the tendency of 

 any such habit or instinct, or for that matter colour or plumage, to 

 appear earlier in the ontogeny than the point at which it developed, 

 a result only restrained, if at these earlier stages it is disadvan- 

 tageous. In the case of the processionary caterpillar there is nothing 

 to prove that the instinct to follow a leader is injurious to the feeding 

 larva, and it is probably even useful, generally, to a larva, who realises, 

 may we say, knows (though the knowing is not exactly of the human, 

 conscious, type of knowing), that a certain region of the tree is stripped 

 of foliage and leads his fellows to new pastures. 



It will of course be evident that a larva, unaware of these circum- 

 stances, unght happen to lead and might lead wrong, as there is no 

 selection of a leader, but only that all follow the one that happens to 

 go first. 



l>ut, specially also, there are comparatively rare, but not perhaps 

 infi'oquent, occasions when a tree is completely denuded, from being 

 small or from having many nests, and it becomes impei*ativc for the 



