REPORT OF THE DIRECTOR IQIG 205 



reasons were given for supposing that the flow of water through 

 the central and older hydrospires was much more copious than 

 the flow through the more lateral and younger hydrospires. 

 That they could so function is now very manifest for there were 

 undoubtedly two water streams, one along each side of the 

 heavier pseudambulacral plates (ten such streams in all) which 

 would serve to augment the flow through the larger hydrospires. 

 Two of these streams near their meeting point received the ejecta 

 of the completely covered anus and swept it through the largest 

 hydrospires of a single deltoid. Even here the volume of 

 oxygen-bearing water must have been in excess of the deoxygen- 

 ated stream from the alimentary canal and the hydrospires in- 

 volved would still in a measure carry on their primitive function 

 of respiration. 



Figure 3 of plate i shows the upper surface of an arm of the 

 type xio. All the covering plates, save six near the end of the 

 arm, have weathered away. Five of these are over the upper 

 (in the figure) row of adambulacrals and the smallest end one 

 over an adambulacral of the lower row. These plates have lost 

 somewhat through weathering but as end and newly-formed 

 plates we should expect them to be small. A portion of one row 

 of adambulacrals is also lost but the outer edges of the remain- 

 ing plates show clearly the openings through which the surplus 

 water drained into the hydrospires. These openings are between 

 the plates but perhaps a little nearer their outer ends than such 

 pores appear to be in the Blastoidea. In fact it seems that in 

 Blastoidocrinus the outer edges of the adambulacra did not meet 

 beyond the opening. The vertical channels with their lateral 

 connections shown in figure 2 were probably covered by a mem- 

 brane possessing ciliary processes, and these accomplished the 

 separation of the food particles and directed them to the food 

 groove. Hambach's beautiful drawing of a portion of the pseud- 

 ambulacrum of a specimen of Pentremites sulcatus, 

 [see fig. 5, plate 2 of his " Revision " ^] shows what was very 

 probably a similar arrangement with the marked diflference that 

 the collecting floor is at right angles to the direction of flow 

 through the pores in Pentremites and parallel with it in Blas- 

 toidocrinus. Each pore in Pentremites is figured las passing into 



1 A Revision of the Blastoidea with a Proposed New Classification and 

 Description of New Species, by G, Hambach, St Louis, Mo. 1903. Nixon- 

 Jones Printing Co. 



