148 DISCOVERY REPORTS 



pp. 155-157). In the region of the brood-pouch, chitinized rods are present in the ventral integument, 

 but there is also a broad unchitinized central area in the integument, which presumably ensures 

 adequate space for the development of the young. The thoracic limbs not involved in the formation 

 of the brood-pouch carry ventral coxal plates and occasionally accessory supporting plates as well. 



(2) Barnard, in the key already referred to, stated that the second ramus of the uropod is absent in 

 the members of the family Idoteidae; this, however, is not the case, for while the absence of this 

 ramus may be a fairly general character among members of the subfamily Idoteinae, it is not universal ; 

 it is present, for example, in Cleantis linearis Dana and C. granulosa (Heller). In the remaining three 

 subfamilies of the Idoteidae, namely, Glyptonotinae, Macrochiridotheinae and Mesidoteinae, its 

 presence, rather than its absence, is characteristic. 



Text-fig. I. Series of diagrams to show the change of position 

 of the penis, (a) Glyptonotus antarcticiis, x 2. {b) Edotia 

 oculata, X 13. (c) Edotia oculopetiolata, x 17. (d) Pseudidothea 

 bonnieri, x 13. p.p. position of penis. 



Text-fig. 2. Idotea indica S, x 6, ventral view. 



(3) A further point must be mentioned in connexion with Barnard's paper, namely, his discussion 

 of the position of the openings of the vasa deferentia in the male. The normal position of these 

 openings in members of the Isopoda is on the ventral surface of the last thoracic somite, either about 

 the middle of the somite, or on its posterior margin. According to Barnard (1920, p. 380), in the 

 members of the Valvifera, the openings have shifted on to the first pleon segment; this certainly seems 

 to be the case in members of the Asticillidae, Pseudidotheidae and Amesopodidae, where the pair of 

 penial filaments are fused to form a single process which, in members of the Pseudidotheidae, are 

 distally cleft (Text-fig. i, d). But the condition in the family Idoteidae as a whole does not agree with 

 Barnard's statement, for the position of the penial filaments is variable, though it is always posterior 

 to the ventral coxal plates of the last thoracic somite. In Glyptonotus antarcticus (Text-fig. i a), for 

 example, the penial filaments spring from the articular membrane immediately behind the fused 

 coxal plates of the last thoracic somite; there is, in this species, a slight excavation of the plates in 

 which the bases of the filaments are lodged. 



(4) Barnard also made a point of the fact that the penial filaments are separate in the Idoteidae and 

 united into a single process in the Astacillidae. This, however, is not true of all the members of the 

 Idoteidae, for in the species of Edotia the basal portions of the two filaments are fused together 

 (Text-fig. ib,c). 



