procession of Gnethocampa pinivora. 577 



with many pines, descending to the entrance of the actual forest. 

 Mr Edwards informs me that most of the processions and the 

 longest individual ones he found were in the Allee de Turenne 

 in the much more favourable weather he experienced. The 

 comparative frequency of processions in this inhabited road is 

 curious, as the pines in the villa gardens are nearly all high and 

 nests seem relatively scarce, while they are very numerous in the 

 large plantation of saplings, averaging 12 ft. high, surrounded by 

 old pines destitute of nests a short distance within the forest. 

 The accompanying photographs show nests in this plantation. 

 I examined a large number of the nests on the saplings and did 

 not find one which contained more than 16 living larvae, and 

 this number was exceptionally high. As there Avas no evidence 

 that the nests were temporarily deserted by larvae in procession 

 it seemed that their occupants must have buried for pupation 

 before my visit, which on the whole agrees with Edwards a fortnight 

 earlier finding many more processions than I did. But if the forest 

 larvae had buried themselves it appears probable that some con- 

 dition holds for the Allee de Turenne which causes certain or all 

 the events of the life cycle to fall later than in the forest proper. 

 The longest processions I found in the Allee were 33 and 68 larvae 

 respectively. These two were close together and there were reasons 

 for thinking that they were a procession of 101 broken. Edwards 

 records one of about 260 larvae, and Fabre observed one of 300. 



As stated above I found no processions in the pine forest 

 itself, and as prolonged observations in the Allee de Turenne were 

 difficult from the traffic I made collections of larvae there and 

 placed them on the sand in the forest. This treatment did not 

 interfere at all with the formation of processions, and when the 

 larvae were placed on a table indoors processions were made as 

 quickly as on the sand in the forest: thus in one case a group 

 of 7 larvae placed on a tablecloth were in regular march in 

 8 minutes. 



It will be noted that the circumstances caused my obser- 

 vations this year to be made under less natural conditions than 

 those by Edwards. Again, some of them were made on members 

 of single processions and others on larvae of different proces- 

 sions mixed together. Thus some of the features noticed may 

 have been due to the larvae being from different nests and of 

 different ages, but I did not find any distinction between the 

 behaviour of what may be called "pure" and "mixed" processions. 

 Larvae certainly from different nests formed processions as readily 

 as those belonging to one procession. But the great tendency of 

 processions to break up into daughter processions recorded by all 

 observers and the readiness with which stray larvae will join a 

 procession render it impossible to regard any procession seen first 

 in march as belonging to one nest. This raises the important 



