sEA-uncinxs. 445 



find a place among the Holectypoida, while EchAnoconus was 

 banished to the Ecliinoneid^e among the Spatangoida. A similar 

 fate befell AnoriJwpijgus, while (Jonoclyjieus, a genus till then 

 iisually classed witli either Echinanthus or Clypeaster, wanhvought 

 into the Holectypoid group. (^C'oiiocl2/peus was regarded as a 

 '' Gcderites" by Grateloup, 5.) 



Duncan divided the Holectypoida (whose brief diagnosis was 

 " Exocyclic, oligoporous Ectobranchiata ") into two unnamed 

 sections. The subdivision was made on the details of the 

 perignathic girdle, and Discoldea and Conocli/peus, on account of 

 the supposed rudimentary state of their processes, were thiis 

 separated from Holectypiis and Fygaster. Galeropygus and 

 Pacliyclypeas were regarded as being Holectypoids, but 

 as not sufficiently known to be definitely associated with, or 

 separated from, any of the other geneiu. 



Duncan's classification was followed absolutely by Sladen in 

 the " Zittel-Eastman " Text Book of Palfeontology (53). 



The only remaining classification of the group in which any 

 important changes are niade is that devised by (iregory in 1896 

 (50), and published in Lankester's ' Treatise on Zoology.' Here 

 the " Holectypina" (a group cori-esponding in part with the 

 Holectypoida) are regarded as a suborder of the Gnathostomata. 

 The chief contrast between Gregory's group and that of Duncan, 

 is that the former author so modifies the diagnosis of the 

 Holectypina as to admit Galerites {Echhioconus), althovigh it is 

 believed to be edentulous. The Holectypina are divided into 

 four families, the Pygasteridse, the Discoidiidje, the Galeritidfe, and 

 the Conoclypeidi^. Discounting the genera desci'ibed since 1889, 

 the Pygasteridpe correspond to section I. of the Holectypoida, 

 with the queried inclusion of Galeropygus and Pachyclypeus. The 

 Discoidiidje and Conoclypeidie together contain the members of 

 Duncan's second section, while the Galeritidse are the first sub- 

 family of the Echinoneidse of the Revision. 



The classification proposed by Gregoiy seems to accord better 

 with our knowledge than any of those previously suggested. In 

 view of the com})lexity of the I'elations of the Holectypoida, owing 

 to its primitive and annectant character, I prefer to regard it as 

 a gi'oup so much apart from the other Irregularia as to merit 

 its retention as an Order, as Diincan originally considered it. 

 After a thorough study of the comparative morphology of most 

 of the genera included in the order, I have attempted to revise 

 its internal classification in such a manner as to indicate the 

 affinities, both internal and external, which the study has made 

 manifest. 



III. The Revised* Classification. 



The characters of an annectant group are inevitably plastic 

 and unstable. For this reason a natural classification of such a 



