54 Psyche [June 



"g?" in Fig. 18, since it may correspond to the structure labeled 

 "pu," instead; but I have provisionally adopted the interpreta- 

 tion indicated by the labeling. In most of the Paiior pa-group 

 there are one or two pairs of dorsal valves (dorsovalvae) , "dv" of 

 Figs. 21, 23, 24, etc., and a pair of ventral valvse ( ventrovalvse) , 

 "vv" of Figs. 23, 26, etc., and it is possible that certain of these 

 valvse may represent the penis hooks of Neuroptera, etc. 



The anal tubercle "ap" of Fig. 21, bears at its base a pair of 

 appendages "c" whose location suggests that they are homologous 

 with the so-called cerci "c" at the base of the median terminal 

 structure "sa" of Figs. 18 and 22. On this account, I would 

 consider the structure "ap" of Fig. 21 as homologous with the 

 structure "sa?" of Figs. 18 and 22, although I am not certain 

 whether the structure labeled "sa?" in Figs. 18 and 22 is the epi- 

 proct "sa," or the proctiger "ap," of other insects. Tillyard 

 describes a pair of segmented cerci in Nannochorista (Fig. 28, 

 "c"), which appear to be homologous with the structures labeled 

 "c" in Figs. 21, 23, etc., and on this account I have interpreted the 

 latter structures as the cerci. I am not certain of the correctness 

 of my interpretation of the structures labeled "c," as the cerci, 

 and the structures labeled "dv," as the dorsal valvse, in Fig. 24 

 of Merope; but have provisionally adopted this method of homol- 

 ogizing them. The projecting ventral process "co" of Fig. 26, 

 may be homologous Avith the columna, and if the latter is repre- 

 sented by the coiled filament "sf " of Figs. 18 and 22, the structure 

 labeled "co" in Fig. 26 is doubtless to be homologized with the 

 coiled filament "sf " also. 



The phallus "pe" is large and prominent in Boreus (Fig. 20), 

 and the hypandrium "hy" is well developed in this insect. In 

 Panorpodes (Fig. 27) the structure which is here interpreted as 

 the hypandrium "hy" shows a marked tendency to become long 

 drawn out and furcate, although the cleft at its apex is not very 

 deep. In the Panorpid shown in Fig. 26, however, the hypandrium 

 "hy" is deeply cleft, and the two arms of the fork are compara- 

 tively long and narrow. The character of the hypandrium fork, 

 the valvse, etc., should be as valuable features for the purpose of 

 classification as any structures, and it is surprising that they are 

 not more employed in taxonomic keys. 



The gonopods "g" of the Mecoptera here figured are composed 

 of two segments. The basal one "pa?" of Fig. 23 may possibly 



