130 ILLINOIS BIOLOGICAL MONOGRAPHS [448 



the Pamphiliidae had previously evolved. The specialization is indicated 

 by the vestigial condition of the ocellarae, subanal appendages, and 

 thoracic legs, on the one hand, and the development of suranal processes 

 on the other. The presence of the functional metaspiracles is of phy- 

 logenetic importance. So far as the head characters are concerned, this 

 family resembles the Tenthredinidae to a limited extent, and in some of 

 the generalized genera of the latter the thoracic legs and the caudal portion 

 of the tenth abdominal segment undergo some modifications which in a 

 remote sense simulate the condition in the Cephidae. But since this 

 family differs from the Tenthredinidae, and resembles the Pamphiliidae in 

 the absence of larvapods and the presence of subanal appendages, it is 

 considered more reasonable to ascribe to it a closer relationship to the 

 Pamphiliidae than to the Tenthredinidae. 



The Xiphydriidae, with its somewhat modified mouth-parts, three- 

 segmented antennae, fleshly thoracic legs, suranal process, absence of 

 larvapods, the general shape of the body, and its biology, resembles the 

 Cephidae but differs from it in theabsence of subanal appendages, ocellarae, 

 and in the vestigial functionless metaspiracles. The absence of the subanal 

 appendages may point to one of the two possibilities in regard to the origin 

 of the Xiphydriidae. This family might have evolved from the cephid- 

 like ancester but have lost the subanal appendages by the completion of 

 the process of atrophy which had already reduced the original distinctly- 

 segmented appendages (similar to those of the Pamphiliidae) to the 

 vestigial papilliform appendages of the Cephidae. The two families 

 under consideration might, on the other hand, have had a common stem 

 which possessed subanal appendages, ocellarae, and vestigial metaspiracles. 

 In the absence of positive support for the first possibility, it is more 

 expedient to consider the second possibihty as nearer to the true rela- 

 tionship of the two families, Xiphydriidae and Cephidae. 



The Siricidae, with its greatly reduced thoracic legs and mouth-parts 

 together with certain other characters, is considered more specialized 

 than the Xiphydriidae. The presence of the functional metaspiracles 

 and its genetic significance have already been discussed. For the same 

 reason which suggests a common origin for the Cephidae and Xiphy- 

 driidae, the Siricidae is considered to have arisen from a common stock 

 which gave rise also to the Xiphydriidae. In the degree of specialization 

 by reduction as well as by addition, this family outranks the Xiphydriidae. 



The Oryssidae, with its vestigial mouth-parts, absence of ocellarae, 

 thoracic and abdominal legs, subanal appendages, suranal process, caudal 

 protuberances, and functional metaspiracles, together with its parasitic 

 habit, is unquestionably the most highly specialized family of the Ten- 

 thredinoidea. Its morphological and biological characters are so different 

 from other families that it is not easy to ascertain the systematic position 



