128 ILLINOIS BIOLOGICAL MONOGRAPHS [446 



an adaptive structure which has arisen in response to the habit of the 

 larvae and does not represent the true appendages of the segment, to 

 which the suranal lobe belongs. For this reason the caudal process is of 

 less significance phylogenetically than the subanal appendages of the 

 Pamphiliidae and Cephidae. 



The metathoracic spiracles of the larvae are either obsolete or vestigial 

 in the majority of the Tenthredinoidea. The larvae of the Cephidae and 

 Siricidae differ from all others in that the metaspiracles are functional 

 and as large as the abdominal spiracles. It is important to ascertain the 

 original condition of the metaspiracles in these families because upon the 

 interpretation of their primitive condition depends their phylogenetic 

 value and hence the relationship between these two families and also 

 between them and other families. It is considered reasonable to assume 

 that the progenitor of insects and hence the ancestor of the Hymenoptera 

 possessed functional spiracles on all the segments of the body including the 

 metathorax, and that their metaspiracles must have been as large as the 

 abdominal spiracles. The closed minute functionless metaspiracles found 

 in the Pamphiliidae, Xyelidae, and others, indicate a condition of atrophy 

 rather than a rudimentary condition, and so far as this character is con- 

 cerned the Cephidae and Siricidae represent the unmodified primitive 

 condition and some sort of relation between these two families must be 

 assumed. But on the basis of other characters it is not conceivable that 

 these two families evolved one from the other in a linear sequence, apart 

 from and independent of other families; they must have descended from 

 a common stock which also gave rise to other families which exhibit 

 vestigial metaspiracles. If this is true there must have taken place a 

 series of dichotomies starting with functional metaspiracles, one line of 

 development resulting in the loss of this primitive character and the other 

 line of evolution retaining the original condition. By assuming four such 

 successive dichotomies in the line of evolution, the origin and significance 

 of the metaspiracles of the Cephidae and Siricidae can be reasonably 

 explained. At each of the four successive dichotomous divisions which 

 produced respectively the pamphiliid-like progenitor and Xyelidae, 

 Pamphiliidae and the cephid-like progenitor, Cephidae and the xiphydriid- 

 like progenitor, and Xiphydriidae and Siricidae, one line of descent always 

 carried the original character and the other line lost it until this peculiarity 

 was generally sifted out, being retained unmodified only in the Cephidae 

 and Siricidae. In this way the metaspiracles are here considered to be 

 the direct descendant of the primitive structures which remained unmod- 

 ified thruout the course of evolution of these families. The two families 

 are generalized in this respect indicating a close genetic relation. 



