NO. 6 INSECT ABDOMEN — SNODGRASS lO/ 



be produced into two or more soft lobes (anl). It is claimed by Keni- 

 ner (1918), from embryological evidence, that the so-called pygo- 

 podium is the rudimentary eleventh abdominal somite. The eversible 

 pygopodial lobes, however, in no way represent appendicular organs, 

 Kemner asserts, nor are they evaginations of the wall of the rectum, 

 as they have been supposed to be ; they are merely productions of the 

 cercumanal area on which the longitudinal muscles from the tenth 

 to the eleventh segment are attached. If the organs in question, there- 

 fore, are not of an appendicular nature, the term " pygopodia " should 

 not be given to them, since it is convenient to apply this name spe- 

 cifically to the true appendages of the tenth, or pygidial, segment. 

 The presence of a distinct though rudimentary eleventh segment 

 in larvae of Coleoptera is of interest because of the general suppres- 

 sion of this segment in holometabolous insects. 



TERMINAL LOBES OF THE PARAPROCTS 



In a few of the lower Pterygota an appendicular lobe is borne by 

 each of the paraprocts. These processes have been termed " para- 

 processi " by Crampton (1920). Examples of paraproct lobes occur 

 in the Odonata and in the tridactylid Orthoptera. 



The paraproct processes of Odonata occur in adult Anisoptera in 

 the form of small, seta-bearing lobes projecting posteriorly from the 

 ends of the paraprocts. Corresponding lobes are not present in the 

 larvae of this group of dragonflies, in which the paraprocts, together 

 with the elongate epiproct, form the valves that close the anal open- 

 ing (fig. 12 A, B). In the larvae of Zygoptera, however, paraproct 

 lobes are highly developed as the large, flat, tracheated plates that form 

 the lateral caudal gills (C, paptl). The median gill (cf) is a similar 

 lobe of the epiproct (Eppt), and is evidently comparable with the 

 median caudal filament of Thysanura (fig. 7 A, B, C, cf). 



In the Orthoptera paraproct lobes are well developed in the Tri- 

 dactylidae, where they have the form of long processes resembling 

 the cerci (fig. 45 A, B, paptl). In some species of RIpipteryx those of 

 the male are incurved at the ends and are said to be used as claspers 

 during copulation. In Ellipes (fig. 45 A, B) each process is borne 

 on a membranous area at the end of the short paraproct {Papt). 

 Crampton (1918, 1921) has given considerable significance to the 

 " paraprocessi " of the Tridactylidae, which he regards as homologous 

 with the styli of the preceding segments, and as representatives of 

 the exopodites of crustacean appendages (fig. 45 C, D, cxpd). Most 

 other writers, including Walker (1919), regard the processes as sec- 



