164 Dr. F. A. Bather — Studies in Edrioasteroidea. 



Tegminals. — In both species, where the grooves meet at the 

 peristome the cover-plates continue round them, and so form a roof 

 over the actinal centre. No special plates are developed here, but all 

 are serially homologous with the other cover-plates and have, in 

 ^. bigshyi, the same accessory plates. There are, however, two 

 differences: (1) in order to cover the space, these plates are rather 

 larger and less regular in shape than normal cover-plates ; (2) they 

 appear suturally united to form a solid tegmen, and in consequence 

 are rarely pressed into the mouth-cavity as in B, but are frequently 

 preserved in place even when the other cover-plates have disappeared 

 (PI. X, Fig. 1). 



The Peristome, which lies beneath this tegmen, consists of 

 a roughly circular opening, surrounded by a frame. Owing to the 

 persistence of the tegminal plates, the peristomial structures are 

 rarely well shown, but it has proved possible to develop them on the 

 right side of specimen A (Text-fig. 3). Here the floor-plates of one 

 side of the right posterior ray curve round the right posterior 

 interradius and meet those of the adjacent side of the right anterior 



median line cover-plates 



III 



Fig. 2. Cover-plates of Edrioaster levis, from the adoral face near the 

 periphery of the holotype. x 6 diam. 



II. From the left anterior ray, showing the curvature of the admedian 

 tract and its gradual change into a distinct plate. 



III. From the anterior ray, showing small plates similar to those in II, 

 but with yet smaller plates between them. 



ray, so that there is a fan-like arrangement of the pore-bearing 

 sutures about the interradius. Eour of these plates, two from each 

 ray, appear to be fused at their marginal ends; and the solid plate 

 thus formed stretches centrifugally along the interradius, serving 

 as a fixed abutment for the adjoining floor-plates to the number of 

 2J on each side. In other specimens five floor-plates may compose 

 this interradial plate. Rarely, as in E 16172, there are two such 

 compound plates formed in an interradius. These plates were noticed 

 by E. Billings (1854, p. 273), who doubted whether their pores 

 penetrated to the interior, but observed and figured a pore in the 

 posterior angle as larger than the others. 



As the floor-plates curve round to meet those of the next radius, 

 they become separated from the corresponding floor-plates on the 

 opposite side of their own radius. Between the two rows on the 

 floor of the groove, there seems to be intercalated another skeletal 



