BY R. J. TILL YARD. 619 



that no close relationsliip exists between the Perlaria and the 

 Odonata 



Secondly, as regards the Plectoptera. The close resemblance 

 in the arrangement of the internal structures of the caudal pro 

 cesses, as compared with those of the caudal gills of Zygoptera, 

 is indeed a striking one, since the number and position of the 

 main nerves, and the positions of the primary afferent blood- 

 canals in the three processes are the same in both. 1"he only 

 differences are tlie long, many-jointed condition of the filaments 

 in Plectoptera, and the absence of a main longitudinal trachea 

 (assuming that this character holds for all Plectoptera). We 

 are, I think, bound to conclude that the relationship between 

 the Plectoptera and Odonata is a more definite one than that 

 between the Perlaria and Odonata, and that, in particular, the 

 Zygopterid gill lias been derived from a caudal process closely 

 similar to that of the Plectoptera, by reduction from a many- 

 jointed to a one- or two-jointed condition. 



(2) Steps in the evolution of the Zygopterid Gill. 



If we examine in section any simple evagination of the body- 

 wall of an Insect Larva, forming a short process — such as, for 

 instance, the cercoids of the Odonate Larva — we shall find that 

 they show only cuticle, hypodermis, and internal Cu 

 bl()od-space(Text-fig.44). No longitudinal nerves, 

 tracheae, or closed blood-canals are present. Thus 

 the first stage in the evolution may be assumed 

 to have been a many-jointed process, long, 

 slender, circular in section, with fairly stout Text-fig 44 * 

 cuticle and hypodeiniis enclosing an undiffer- 

 entiated blood-canal, and without either main nerves-, tracheie, 

 or closed afferent blood-canals. 



The next stage would h>e the development of this organ into 

 an organ of touch, by the outgiowth of longitudinal nerves along 



* T.8. through a cercoid of the well-grown larva of Syvle>*tes weyerM 

 Self's; />/, blood; en, cuticle; hin^ hieniooele; liy, liypoderniis; nn, nucleus 

 uf same; ( x 88,'. 



