14 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I46 



larval forms is the true metamorphosis of the insects. When the 

 larva has served its purpose in the Hfe of its species, it is practically 

 destroyed and the developmental process reverts to the adult, which 

 alone can perpetuate the species. The larval destruction and the 

 adult reconstruction take place simultaneously in the pupa which is 

 itself a preliminary stage of the adult. 



It is thus clear that the apparent change of the larval insect into 

 the imago is not truly a metamorphosis. The term metamorphosis 

 means literally "a change of form." The change from caterpillar to 

 butterfly, however, is a change of form only in the eye of the be- 

 holder. Actually the change is a replacement of the larva by the 

 butterfly. The writer has suggested the term retromorphosis for the 

 reversion of morphogenesis to the adult line of development after 

 the dissolution of the specialized larva (Snodgrass, 1961). The de- 

 velopment of the adult and the destruction of the special larval tis- 

 sues go on at the same time in the pupa, but the result is not a 

 transformation of the larva into the adult. 



Recapitulation: This term as applied to individual development 

 implies that an animal in its ontogeny goes through stages of de- 

 velopment that represent successive adult forms in its phylogenetic 

 history. Garstang (1922) has severely criticized this theory in its 

 general concept, contending that the ontogenetic form in one gen- 

 eration represents the ontogeny of preceding generations. This is 

 particularly true of the larvae of holometabolous insects, which may 

 be wholly adapted in their structure to their own way of living and 

 feeding, and in no way represent adult ancestral forms of their 

 species (see Larva, Metamorphosis). 



The development of the insect embryo also is in many ways an 

 adaptation to its life in an eggshell in which it cannot possibly fol- 

 low the evolution of its free-living ancestors. In minor ways, how- 

 ever, the embryo is not necessarily prohibited from recapitulating 

 the evolution of adult structures. It goes through a polypod stage, 

 for example, when limb rudiments are present on all of the segments. 

 At this stage it evidently represents a centipede-like ancestor. More- 

 over, the growth and development of all the appendages from simple 

 undifferentiated lobes should repeat a similar origin of the limbs in 

 some remote wormlike progenitor of the arthropods. Likewise, the 

 development of wings from simple outgrowths of the integument, 

 whether they remain on the surface or are temporarily sunken into 

 pouches, would appear to repeat the evolutionary development of 

 wings from paranotal lobes. In other words, ontogenetic recapitula- 



