302 CYSTOIDEA, BLASTOIDEA, AND HOLOTHUROIDEA. 



their internal tubes are connected with almost all the plates 

 of the test ; in other cases they are restricted to a few 

 (generally three) definite and limited areas the so-called 

 " pectinated rhombs." In the Blastoidea structures of an 

 essentially similar character are found, but they are now 

 wholly confined to the pseud-ambulacral areas, thus having 

 a connection with the brachial system that is entirely un- 

 known among the Cystideans. If we examine the median 

 groove of one of the pseud-ambulacra of a Pentremite, we 

 find that the floor of this is perforated by a series of minute 

 pores. These pores have been shown by Eofe and Billings 

 to open into flat tubes which have delicate calcareous walls, 

 and are placed within the cavity of the calyx. As demon- 

 strated by transverse sections of the calyx, the precise form 

 and number of these tubes differ in different species, though 

 their general arrangement and structure are the same. In 

 some forms (fig. 177, c) there are two of these flattened 

 respiratory tubes to each pseud-ambulacrum, one on each 

 side of the ambulacral furrow. Each tube is flattened ex- 

 ternally and becomes dilated towards its inner or closed 

 side ; and they are not only connected with the exterior by 

 the minute pores in the ambulacral furrow, but they also 

 open in pairs by apical apertures, which will be immediately 

 described. In other species, though there are really only 

 two tubes to each ambulacral furrow, each of these two 

 becomes compound by a reduplication of its delicate wall, 

 and thus splits up into secondary tubes (fig. 177, D). That 

 the system of pores and internal tubes here described is 

 respiratory in function, and strictly homologous with the 

 system of " hydrospires " or " pectinated rhombs " in the 

 Cystideans cannot be doubted, though there is here the 

 peculiarity of their being directly connected with the 

 recumbent arms. 



The pseud-ambulacral areas have been treated above as 

 if composed of a single piece each ; but in reality each is 

 compound in its nature. Thus the floor of the ambulacral 

 furrows is formed by the so-called " lancet-plates " of Eoemer, 

 which generally, or always, are very narrow bifurcate plates, 

 which start at the centre of the calyx and send a long arm 



