procession of Cnethocampa pinivora. 585 



head in the manner characteristic of all isolated larvae and of 

 any larva not in contact with the tail of another, and thus it 

 checked the progress of the procession. In only one case did the 

 hesitation last for more than two minutes. In all cases the old 

 leader placed in head-and-tail contact with the last larva joined 

 up immediately. This and similar observations made on other 

 groups also suggest how little necessary is the thread for keeping 

 the integrity of a procession. The extent to which the leader for 

 the time being takes the initiative has been discussed by Fabre 

 (loc. cit. p. 337), and Edwards (loc. cit p. 432). 



The influences which determine the direction of a procession 

 on the march remain very obscure. Fabre (pp. 367, 368) states 

 his belief that sunlight is the guide, but adds "les points d'ou 

 vient le plus de chaleur sont les pr^feres." In the case of 

 processions I have seen in the woods the general direction 

 appeared to be towartls the greatest sunlight through the foliage, 

 and the sidewalks and roadway of the Allee de Turenne at 

 Arcachon in which processions are so often seen are more sunlit 

 than the gardens from which they come : again in the case of 

 processions indoors the course taken was usually towards the 

 window. Edwards (loc. cit p. 432) describes how the leader of a 

 procession followed rays reflected artificially on to him alone. 

 This year, April 4th afforded the only favourable opportunity 

 I had of observing the possible effect of sunlight, as that was the 

 one day of sunshine, bright intervals occurring between showers. 

 Two short processions of three larvae each and a single procession 

 resulting from artificial mixing of these certainly did not march 

 towards the sun. The general direction taken was either at right 

 angles to the path of sunlight or away from the light. In one 

 case the sudden breaking out of light from between clouds was 

 at once followed by the procession of six turning away from it. 

 In this and on other occasions attempts to make the processions 

 march towards the sun failed altogether. These processions were 

 on sand clear of trees, but from 10 to 20 feet away from a zone 

 of saplings. The latter were away from the sun and the general 

 direction of the larvae was towards them. Whether smell or other 

 sense attracted the larvae towards the trees cannot be asserted, 

 but they certainly did not follow the path of greatest light. We 

 remain in ignorance of the stimuli which influence the behaviour 

 of a procession. The remarkable dorsal papillae of the larvae 

 may be " appareils de metdorologie " as Fabre has suggested 

 (loc. cit. p. 366), and in various respects it is likely that the larvae 

 respond to influences we cannot comprehend. 



Burying for pupation has been described in detail by Fabre (loc. 

 cit. pp. 368, 369) and also by Edwards (loc. cit. p. 435). I watched 

 it frequently this year and found that as a rule all the larvae of a 



