i86 THE BIOLOGY OF INSECTS 



often distinguished as a " campodeiform " larva. As the 

 wingless bristle- tails are the most primitive of living insects, 

 it has been suggested that the campodeiform is the primitive 

 type of insect larva, and that larvae such as caterpillars or 

 chafer-grubs — the *' cruciform " type of A. S. Packard 

 (1898) — are to be regarded as more strongly modified in 

 correlation with their habits, which differ more markedly 

 from those of their adults. In the maggot of a muscoid 

 fly we see a still more profoundly modified, in fact degraded, 

 " vermiform " type of larva. Confirmation of the opinion 

 that these types indicate an increasing degree of divergence 

 between larva and imago, is afforded by those insect life- 

 histories which exhibit more than one larval type in the 

 course of development. For example, the young of many 

 oil and blister beetles are hatched from the egg as tiny 

 active, armoured campodeiform larvae which seek to attach 

 themselves to the body of a bee, and if successful, are carried 

 to the nest, where, after the first moult, they become changed 

 to soft-coated grubs feeding on the stored honey. The 

 campodeiform precedes the cruciform type in the one life- 

 history, and the inference is drawn that the former con- 

 dition is exceptionally retained because the parent beetle 

 cannot enter the bee's nest so as to lay her eggs where the 

 grubs will spend the greater part of their term of existence, 

 and the active long-legged, armoured form of larva is the 

 best adapted for making its way thither. In the case of 

 the vast majority of metamorphic insects which place their 

 eggs upon, within, or adjacent to the material whereon their 

 grubs will feed, the young insect, as soon as hatched, 

 conforms to the cruciform or the vermiform type. 



These considerations throw light on the nature of the 

 primitive immature insect and on the problem of the origin 

 of the more specialised and degraded larval types. It may 

 now be advisable to pass to another problem which con- 

 fronts the student of insect development : the relation of 

 the open (exopterygote) to the hidden (endopterygote) 

 manner of wing- growth. We have seen that the former is 



