I04 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 85 



such as occur on the more anterior segments of the abdomen. The 

 free lobes (Vs) are evidently comparable with the plantae of the ab- 

 dominal legs of caterpillars (fig. 37 D, Vs), though the retractor mus- 

 cles of the lobes in the sawfly larva take their origins from a point 

 on the side of the segment just anterior to the cleft of the anal open- 

 ing (fig. 43 A, rvs). In other forms, as in Cimhex americana (B), 

 the ventral appendages {Pp) consist apparently of the plantar lobes 

 only, which arise directly from the flattened venter of the pygidial 

 segment. Again, as in the Cephidae and Pamphiliidae, typical " pro- 

 legs " are replaced on the tenth segment by slender jointed appendages 

 (C, Pp) ; but these " arthrostyli," as they have been called (Cramp- 

 ton, 1919), are evidently alternative forms of the anal " prolegs," or 

 postpedes, since they arise at approximately the same points as do 

 the typical postpedes in other forms (A, Pp), and do not occur con- 

 jointly with the latter. They are not, however, provided with muscles, 

 so far as the writer could discover ; but each is penetrated by a large 

 nerve, and bears sense organs on the distal segment having the form 

 of minute disks, in addition to setae on the proximal and middle 

 segments. The jointed form of the terminal appendage is, therefore, 

 evidently an adaptation to a sensory function instead of a locomotory 

 one. 



Crampton (1919) suggests that the jointed appendages, or " arthro- 

 styli," of chalastogastrous larvae do not represent the " prolegs " 

 directly, but that each has the relation to the latter of a stylus, that 

 is, it is an appendicular part of the true appendage. Middleton (1921), 

 on the other hand, thinks that the jointed appendages are direct repre- 

 sentatives of the unjointed postpedes, because the two organs have 

 identical positions on the tenth segment, and do not occur together. 

 He would attribute the difference in form to the different habits of the 

 larvae, since those species having typical, fleshy postpedes feed in the 

 open on leaves and grasp the edges of the latter with the terminal 

 appendages, while those having slender, jointed postpedes, bore into 

 the stems of plants, or live in the protection of web nests or curled 

 leaves. To the writer it appears most probable that the two forms 

 of appendages are identical organs, and that the jointed variety is a 

 secondary modification of the typical postpedes for a sensory function, 

 the jointing being a mere subdivision of the appendage and not a true 

 segmentation. 



The observation made by Middleton (1921) that the postpedes 

 of the larva become the cercus-like appendages of the adult sawfly has 

 been discussed in the introductory part of this section, and need be 

 given no further attention here, except to point out its importance, 



