STRONGYLOCENTROTID.^. 341 



since it was not included in the genus by the original describer of that group. 

 It is possible that Sphserechinus may still be used for some fossil forms but con- 

 sidering S. roseus Russo a synonym of granulans and placing the latter in Strongy- 

 locentrotus, Sphserechinus is not available for recent Echini. The differences 

 between the pedicellariae of granulans and those of its allies in Strongylocentrotus, 

 upon which Mortensen lays much weight, seem too trivial to be of any real 

 value. So too with regard to the characters by which Pseudocentrotus and 

 Anthocidaris are supposed to be distinguished. Had Dr. Mortensen examined 

 larger series of specimens, it is improbable that he would have attempted to 

 isolate those two genera. 



There are two species, regarding whose status Mortensen expresses his 

 doubt owing to his inability to examine sufficient material, and of which good 

 series are at hand. One of these, Sphcerechinus ausiralice A. Ag., becomes, as 

 suggested above, the type of the new genus Pachycentrotus. The other, Tox- 

 opneustes gibhosus Agas. and Des., while related to Paracentrotus in many ways 

 has such a modified abactinal system and such specialized globiferous pedicel- 

 lariae, that it is better to make it the type of a new genus, Caenocentrotus. WTiile 

 therefore rejecting three of Mortensen's genera, two new ones are here proposed 

 and thus the family contains only one less genus than he suggested. 



The eight genera of Strongylocentrotidae adopted, are distinguished from 

 each other as follows; but the arrangement is very artificial and the various 

 characters are used arbitrarily regardless of their real importance, the only 

 object being to make the accepted genera tangible and easily recognized. 



Test with ambitus above equator; not more than 4 pore-pairs in each arc; ocular 



plates all fully exsert; primary spines longest on flat abactinal surface . . Echinostrephus. 

 Test with ambitus at or below equator (if above, there are more than 4 pore-pairs in 

 an arc, and one, at least, of the oculars is insert). 

 Test thin with deep gill-cuts, having a very prominent flange ("lip" or "tag") 

 on interradial side; plates of buccal membrane carry both spines and 



pedicellariaj Pseudoboletia. 



Test seldom thin; gill-cuts without a prominent flange, and usually shallow; 

 plates of buccal membrane do not carry spines. 

 Pore-pairs in arcs of 4-10; if more than 5, valves of globiferous pedicellariae 

 with one or more lateral teeth. 

 Usually all oculars exsert but sometimes one and rarely more, are 

 insert; buccal membrane with scattered plates; valves of globif- 

 erous pedicellariai with one or more lateral teeth on each side. 

 Test not very stout; primary spines usually rather long, at least 

 .25 h. d.; valves of tridentate pedicellaria; not pecuhar; pore- 

 pairs 4 or 5 Paracentrotus. 



Test very stout; primary spines rather short; valves of tridentate 



