145 



growths which first arise after the last traces of the abdominal 

 limbs have disappeared, and in a more median position than the 

 latter occupied. The first traces of gonapophyses, contrary to 

 earlier statements, were found in very young larvae or n\'mphs 

 of Agrión pulchellum of but 2 mm. abdominal length. These 

 traces are referred to the lateral gonapophyses of the ninth seg- 

 ment, and characterise both sexes. In larvae of 3 mm. abdominal 

 length and more, the sexes can be distinguished owing to the 

 appearance of the first rudiments of the median gonapophyses 

 of the same segment in the female but not in the male, soon 

 followed by the rudiments of the anterior gonapophyses of the 

 eighth segment in the former sex. He found rudiments of the 

 gonapophyses in quite small ^Eshnid larvae, but not in those of 

 Gomphidae, Cordulegastridae, CorduHdae, and Libellulidœ. While 

 all three pairs of gonapoph3^ses are present in the imagos of the 

 Zygoptera, of the .-Eshnidae and of the Petaluridic and form an 

 ovipositor, reduction in size in the lateral gonapoph^'scs is ap- 

 parent in the latter two groups and becomes almost absent in 

 the Cordulegastridai. In the remaining Anisoptera the lateral 

 and the median gonapophyses are almost, or altogether, absent, 

 and the anterior pair show all degrees of reduction and fusion. 

 This reduction of the gonapophyses coincides with van der 

 Weele's idea of the higher differentiation of the Anisoptera. 

 The genital pore arises originally in both sexes behind the middle 

 of the ninth sternite, and persists in this position in the male. 

 In the female it is shifted forward so that it has been described 

 as at the hind end of the eighth segment. 



A reduction of the female gonapophyses within a more 

 limited group {Synihemis — Corduliinae) has also been demon- 

 strated by TiLLYARD (1910), who has emphasised (1909) the 

 correlation between elongated eggs and the presence of an ovi- 

 positor on the one hand, and the wider, less elongated eggs of 

 those Odonata in which the female lacks such an organ, on the 

 other. 



Wing Venation. 



Although the general characteristics of the wings were em- 

 ployed by LiNN.EUS and his successors in defining the various 



19 



