624 CAUDAL GILLS OF ZYGOPTERID LARV^, 



Turning next t() tlie Ayrionidiv, we Ijave a very distinct prob- 

 lem to deal with. 1 think that the evidence liere is all in favour 

 of the development of the Agrionid form of Vertical Lamellate 

 Gill-type direct from a two-jointed or constricted Saccoid Type, 

 such as still exists in Neosticta, without the intervention of a 

 Triquetroquadratp, Type at all. The reasons for this are as 

 follows : - 



(a) Neosticia can hold its gills fairly widely apart, and they 

 are not swollen out so greatly (as in the case of Diphlebla) that 

 tlie lateral gills would be compelled, in time, to take on a triquetral 

 form, by pressure of the median gill from above and between 

 them. Consequently, there is an opportunity for the develop- 

 ment of a quadrate form in all three gills. 



(b) The gill being two-jointed, such a quadrate form naturally 

 suggests itself as suitable for the basal joint or stalk of the gill. 



(c) All three gills of the Agrionid form of Vertical Lamellar 

 Gill-Type do actually show this quadrate forvi in transverse sec- 

 tion. (See Text-fig. 22 in Part ii. of this paper, p. 90). 



{d) The internal arrangement of parts in these gills is that of 

 the Saccoid Type, except only that the median gill possesses tiro 

 main trachea?, instead of four. That this is due to reduction is 

 shown by the fact that regenerated median gills of the Agrionid 

 form of Vertical Lamellar Gill-type show an increase in the 

 number of main trachejt^ to four, as well as a distinct tendency 

 to a narrow Saccoid form. (N.J3. — Regenerated lateral gills also 

 show an increase to three or four tracheie, but three main trachea? 

 are also quite frequently met with in Neosficta, as well as in 

 other Saccoid Types). 



The Constricted or two-jointed Ivamella of Jsosticta is clearly 

 a direct development from the Constricted Saccus of Neoaticla. 

 The two genera are too closely allied to admit of any doubt on 

 that point. 



The larva of Caiiagrion biUvnyhurf-ti, when half-grown, (pro- 

 bably the tenth or eleventh instar) shows the constricted condi- 

 tion almost as completely as in Iso»ticta. Thus the iSodate form 

 finally attained in this and other Agrionine genera must be 

 derived from the older Constricted form, as seen in Jsosticta. 



