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



of that ray, is " a peculiar margined depression ".. If, as Dr. Eoerste 

 suggests, "a duct passed by this path," then the duct in question 

 would most naturally be the hydropore-canal. 



The Relations of the Water- vascular System are of much 

 morphological importance. The most natural interpretation of the 

 appearances in the Edrioasteridae is that the perradial water-vessels 

 lay at the bottom of the subvective grooves, on the ventral side of the 

 abutting floor-plates, and that each gave off alternate branches to 

 podia placed near the outer margin, and that each branch was also 

 connected, through the pore between adjacent floor-plates, with an 

 ampulla lying below (dorsal to) the subvective skeleton. This inter- 

 pretation will be confirmed by comparison with a modern starfish. 

 The perradial vessels passed to the peristome, and JSdrioaster buchianus 

 has already yielded evidence (Study II, 1900, p. 198) to show that 

 they were there connected to form a hydrocircus surrounding the 

 opening to the gullet, just above the mouth-frame. The hydropore 

 lay in the posterior interradius, close to ray Y, and passed through 

 one or more interradial plates of the theca in a sloping direction from 

 right to left. This direction indicates some torsion of the hydropore- 

 canal and presumably also of the stone-canal, both of which must 

 have been in the thecal cavity. Precisely where they became 

 connected with the hydrocircus we cannot say, but it is plain that 

 connection could readily have been effected by anyone of the ampullar 

 or podial pores, and we may here recall that E. Billings observed 

 a pore in the posterior angle to be larger than the others (Study IV, 

 1914, p. 164). 



In the Agelacrinidae there is no such direct structural evidence for 

 the position or even for the existence of perradial canals, not to 

 mention podia. The shape of the cover-plates in most genera does, 

 however, seem consistent with (one is tempted to say " calculated 

 for") the extrusion of podia between them, and it seems natural to 

 suppose that these structures were present, though unprovided with 

 intra-thecal ampullae. An external hydropore has not as yet been 

 detected, but it is conceivable that the hydrocircus opened into the 

 oral vestibule, and that it may have been connected with some canal 

 passing up in the posterior interradius {vide supra). 



JEdrioaster, Agelacrinus, and Steganoblastus present three modi- 

 fications of an original subvective skeleton consisting of paired 

 alternating floor-plates and cover-plates. In Agelacrinus, where 

 ampullae were still unformed, the floor-plates became partly fused, 

 but were still separated at intervals so as to maintain some flexibility. 

 In JEdrioaster the development of ampullae and more vigorous podia 

 produced pores and combined with the general flexibility of the test 

 to keep the original floor-plates distinct, except in the mouth-frame. 

 In Steganoblastus the ampullae and podia remained, but the floor- 

 plates fused and, in correlation with the rigidity of the theca, formed 

 a single piece stretching right along under the groove. 



