439] LARVAE OF THE TENTHREDINOIDEA—YUASA 121 



synoptic descriptions and keys. There are still others who believe in the 

 intrinsic importance of the immature stages in the study of phylogeny. 

 The reasonableness of the oft-repeated objection which was voiced by 

 Comstock (1918) that the larvae of insects exhibit a cenogenetic develop- 

 ment and, therefore their ontogeny bears little or no relation to the phylog- 

 eny of the race, must be admitted in regard to certain structures which 

 are entirely too adaptive and too much modified by environmental factors 

 in meeting the trophic requirements of particular species or genera. But 

 admission of this fact is not incompatible with a belief in palingenesis 

 of other structures. Besides, the warning that the cenogenetic peculiari- 

 ties, which may be of value as distinguishing characters, are of no phylo- 

 genetic significance and must, therefore, be judiciously and discriminately 

 distinguished from more important palingenetic characters, applies not 

 only to the classification of the larvae but to the taxonomy of the adults as 

 well. This objection alone does not invalidate a belief in the intrinsic 

 importance of the immature stages from the phylogenetic point of view. 

 While the writer does not minimize the danger of a too confident expecta- 

 tion of finding phylogenetic indices in the successive ontogenetic stages in 

 entometabolous insects, yet he is equally reluctant to abandon his hope in 

 regard to the taxonomic value of the characters of immature insects. 

 The present study is a partial justification of his contention. 



Students of the Tenthredinoidea have recognized the practical im- 

 portance of the larvae in determining the systematic position of different 

 taxonomic units. Norton (1867) stated that "Mr. Walsh has shown 

 that in some species of Euura and Nematus bred by him, it was almost 

 impossible to detect any difference in the imago, while the larvae varied 

 greatly. Doubtless our opinion will be greatly modified by future dis- 

 coveries." Cameron (1882) was of the opinion that the larvae were of 

 great value in differentiating the tribes and subtribes altho they appeared 

 to be of little use in regard to the genera. MacGilHvray (1913) goes further 

 and states that "it was hoped from a study of the immature stages of the 

 Tenthredinoidea that some information might be obtained as to the 

 validity of the species based on obscure anatomical details." Rohwer 

 also often uses the characters of the larvae as collateral evidence in decid- 

 ing the systematic position of certain subfamilies and genera. 



Nothing definite is known in regard to the ancestors of the Hymenop- 

 tera beyond the probability that they have somehow arisen from a primi- 

 tive type of some neuropteroid-like Palaeodictyoptera. The order is 

 considered to be one of the most, if not the most, highly specialized of all 

 insects. Systematists are unanimous in regarding the Tenthredinoidea 

 as the most generalized of the Hymenoptera. It is difficult, if not impossi- 

 ble, to conjecture the primitive larval type of the Tenthredinoidea. 

 Judging, however, from what are universally considered to be generalized 



