WATEIt SYSTEM. 099 



immediate vicinity of the actinostome the plates become smaller again, the 

 ambnlacral pores increase in distance to form the buccal rosette, which 

 has received the name of phyllode from Desor, and was supposed to be 

 eminently characteristic of the Cassiduloids alone, but which evidently is 

 also found in the Spatangoids proper. In the Clypeastroids the innumerable 

 ambulacra] pores, found both in the horizontal sutures above the ambitus 

 and in the ambulacal furrows of the actinal side, often extending into the 

 interambulacral spaces, give the ambnlacral system a widely different 

 structure from that of the Spatangoids. The buccal rosette is limited to 

 its simplest expression in Echinarachnius ; PL XXXI. /'. / ;, /.;, shows the 

 simple terminal tentacles immediately adjoining the actinostome. 



In the Spatangoids the buccal tentacles have usually been considered as 

 gills ; but from what Mr. Robertson has shown in the case of Echinocardium. 

 they are evidently to be considered as organs of locomotion and prehension, 

 and seem, in spite of their apparent delicacy of structure, capable of exert- 

 ing considerable power in digging into the ground and enabling the helpless 

 Spatangoids to fill their alimentary canal with the fine sand from which 

 they derive their main supply of food. To assist them in this operation 

 the actinal membrane is found to be quite movable, capable of consider- 

 able expansion far beyond the line of the lips of the actinostome proper. 

 The buccal tentacles terminate at the extremity in a cluster of irregularly 

 arranged club-shaped appendages, strengthened internally by limestone 

 rods {PI. XXXII f. 16). This is a totally different structure from that of 

 the so-called gills of the regular Echini (compare PL V.f. .:. where I have 

 figured the brandling gills of Arbacia). These grape-like gills are not con- 

 nected with the circular canal but with the interior cavity ; they are there- 

 fore filled by the introduction of water through the edge of the madre- 

 poric body. Similar gills are found in all Desmosticha, although Midler 

 denies their presence in Cidaridae ; even there they are found, but close 

 to the jaws. The buccal tentacles of the Spatangoids and Clypeastroids 

 are the homologue of the ten large buccal tentacles (PL V.f. ;), placed 

 immediately around the mouth in the regular Echini, and which pierce 

 the ten large prominent actinal plates found on the actinal membrane of 

 all Desmosticha, whether that membrane be regularly imbricated or not. 



In the Clypeastroids proper we have seen that these buccal tentacles differ 

 in no wise from the adjoining ones of the ambnlacral system, and in some 

 Spatangoids (Khynchopygus) the buccal tentacles are much simpler {PL 



