34 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I23 



other insects. Dorsally at the sides of the median lobe is a pair of 

 smaller appendages (Cer) that have the exact relative position of the 

 usual abdominal cerci. However, most students of Odonata, follow- 

 ing Heymons (1904), regard the lateral anal lobes as the true cerci 

 and the dorsal cercuslike appendages as secondary structures, which 

 are termed "cercoids." The reason for this interpretation as given by 

 Heymons is that the "cercoids" are not present in early larval stages, 

 and that the lateral anal lobes in some young zygopterous larvae (E) 

 are long filaments resembling the filamentous cerci of Thysanura and 

 larvae of mayflies and stoneflies. Neither of these facts, however, is 

 necessarily evidence that the "cercoids" are not the cerci of odonate 

 lar^'ae. 



At the transformation of the male anisopterous larva to the adult, 

 the "cercoids" become the superior claspers (fig. loE, F, C^r), the 

 dorsal anal lobe (Eppt) becomes the so-called "inferior" clasper, 

 which arises above the anus (An), and the lateral anal lobes of the 

 larva become small triangular plates lying at the sides of the anus 

 (Papt). It seems incredible that these lateral plates of the adult can 

 be cerci ; they have the exact position of paraprocts relative to the 

 "inferior" clasper and the anus, and there can be no question that the 

 "inferior" clasper is the epiproct. Schmidt (1933) accepts Heymon's 

 interpretation that the dorsal appendages are "cercoids," but he re- 

 gards the paraproctial plates of the adult as sternal plates of the 

 eleventh segment, which they probably are, and he contends that the 

 "true cerci" are small processes arising from the membranous inner 

 surfaces of these plates (fig. 10 H, 711). These processes of the para- 

 procts are shown by Asahina (1949) to be of unusual size, though 

 but little sclerotized, in both the male and the female of the anisozy- 

 gopteron Epiophlebia superstes. Since true cerci in other insects never 

 have any such relation to the sternum as these paraproctial processes 

 have to the paraprocts, this last interpretation is hardly acceptable. 

 On anatomical evidence, therefore, the writer sees no reason for not 

 regarding the "cercoids" as true cerci, and the anal lobes as the epi- 

 proct and the paraprocts in both the larva and the adult. This inter- 

 pretation of the terminal structures of Anisoptera has been maintained 

 also by Crampton (1918) and by Handlirsch (1926). In an adult ani- 

 sopterous female the end of the abdomen (fig. 10 G) is so typically 

 orthopteroid that it might pass for that of a grasshopper or a cricket. 



The apical appendages of zygopterous larvae, whether of the broad 

 lamellar form (fig. 11 A) or of the slender triquetral type (B), are 

 supported on basal plates (A, Eppt, Papt) that evidently are the true 

 epiproct and paraprocts. Each appendage is separated from its sup- 



