V.] METAMORPHOSES OF INSECTS. 95 



before larvae and pupae." This opinion has been 

 adopted by Mr. Packard l in his " Embryological 

 Studies on Hexapodous Insects," 



M. Brauer 2 also considers that the vermiform larva 

 is a more recent type than the Hexapod form, and is 

 to be regarded not as a developmental form, but as 

 an adaptational modification of the earlier active 

 hexapod type. In proof of this he quotes the case of 

 Sitaris. 



Considering, however, the peculiar habits of this 

 genus, to which I have already referred, and also that 

 the vermiform type is altogether lower in organization 

 and less differentiated than the Campodea form, I 

 cannot but regard this case as exceptional ; one in 

 which the development has been, as it were, to use an 

 expression of Fritz Miiller's, " falsified " by the struggle 

 for existence, and which therefore does not truly in- 

 dicate the successive stages of evolution. On the 

 whole, the facts seem to me to point to the con- 

 clusion that, though the grublike larvae of Coleoptera 

 and some other insects, owe their present form mainly 

 to the influence of external circumstances, and par- 

 tially also to atavism, still the Campodea type is 

 itself derived from earlier vermiform ancestors. 

 Nicolas Wagner has shown in the case of a small 

 gnat, allied to Cecidomyia, that even now, in somej^ 

 instances, the vermiform larvae possess the power of 

 reproduction. Such a larva (as, for instance, Fig. 57) 

 very closely resembles some of the Rotatoria, such 

 for instance as Albertia or Notommata, which how- 



1 Mem. Peabody Academy of Science, v. I. No, 3. 



2 Wien. Zool. Bott. Geseils. 1869, p. 310. 



