124 ILLINOIS BIOLOGICAL MONOGRAPHS (442 



leaf-miners have both thoracic and abdominal legs reduced in the number 

 and size of their segments. The small size and reduced number of seg- 

 ments are correlated with the well-developed thoracic legs of the Hylo- 

 tominae. In this subfamily the claws are very large, sharply curved, and 

 provided with empodia-like distal structures, indicating a great adapta- 

 tion for clinging to leaves. This fact is sufficient to account for the reduc- 

 tion of the larvapods. The Pamphiliidae and four specialized families 

 in the same line of development are entirely without larvapods. It is 

 highly desirable to determine whether this apodous condition signifies 

 a common origin of all five families. Upon this question hinges much of 

 the interpretation of the phylogeny of the Tenthredinoidea. 



In characterizing the larvae of the hypothetical primodial Ten- 

 thredinoidea the abdomen was considered as provided with the maximum 

 number of appendages, including ten pairs of larvapods and a pair of 

 subanal appendages. This assumption is based upon the fundamental 

 fact that the progenitor of insects having evolved from a typical arthro- 

 podan organism possessed the typically arthropodan character, abdominal 

 appendages. This assumption is justifiable in view of the following facts: 

 (1) the possession of appendages on all of the abdominal segments is a 

 fundamental arthropodan characteristic; (2) the embryos of practically all 

 insects exhibit at some time during their development rudiments of 

 abdominal appendages; (3) appendages are present on all or some of the 

 abdominal segments in the postembryonic stages of the Apterygota; 

 (4) the gonapophyses of the Exometabola represent the true abdominal 

 appendages in this group of insects; (5) the larvapods and other appenda- 

 ges are present in the larvae of the Mecoptera, Lepidoptera, generalized 

 Hymenoptera, and, possibly, in some other orders, — all these facts indi- 

 cating the wide occurrence and fundamental continuity of abdominal 

 appendages in the Hexapoda. It is, therefore, not unreasonable to assume 

 that the progenitor of insects, at least in some stage of its development, 

 possessed appendages on all of the abdominal segments. The same 

 argument supports the contention that the ancestors of the Hymenoptera 

 undoubtedly closely resembled the remoter ancestors of the Insecta. 

 The larvae of the progenitor of the Hymenoptera for this reason have 

 been considered as provided with the maximum number, or ten pairs, 

 of larvapods, a pair on each of the first ten abdominal segments and a 

 pair of subanal appendages on the eleventh abdominal segment. If this 

 assumption is true, the larvae of the Xyelidae, which possess ten pairs of 

 larvapods, must be considered as representing the most primitive condition 

 found in the Hymenoptera. Graber (1890) has shown that in the larvae 

 of Hylotoma the larvapods arise from the embryonic limb-rudiments and 

 are directly evolved from them during the development and, therefore, 

 the larvapods are the true appendages of the abdomen, homodynamous 



