6 PEOr. p. M. DtJKCAN S EEVISION OF THE 



thuridse. The descriptions of some genera by Meek and "Worthen 

 and McCoy require additions, in consequence of this work of 

 A. Agassiz. 



The foundation of a classification upon the presence or absence 

 of primary tubercles upon tests was due to McCoy, and it served 

 its time ; but it was too artificial and permitted genera to be 

 closely associated which had structural difi'erences of great phy- 

 siological importance. Por instance, some genera with two and 

 with more than two vertical rows of ambulacral plates were 

 associated closely ; moreover, genera which had bevelled plate- 

 edges, and where there was overlap, were associated with forms 

 , which had rigid tests and no overlap of plates. It must be 

 admitted that too much has been made of the presumed and 

 real imbrication of plates in classification. 



Whilst it is undeniable that our knowledge of the genera of 

 Palseechinoidea has increased, it is still necessary to remember 

 that any classification must be open to exception. It is not 

 possible to place some of the genera in the same groups upon 

 the admitted principle of the preponderating taxonomic value of 

 the structure of the ambulacra, and, indeed, there must be some 

 anomalous genera ; but the following grouping is suggested as 

 the best at the present time. 



All genera founded in reference to single plates or spines are 

 not considered, and JEocidaris, Keyserl., is omitted, for it is a true 

 Cidarid. Palceocidaris, Beyr., is synonymous with Lepidocentrus, 

 J . Miiller. FeriscJiocidaris, JSTeum., is the same as Perischodomus, 

 McCoy. Echinocrinus, Ag., and 'Palceocidaris, Desor, are 

 synonymous with Archceocidaris, McCoy. Protoechinus, Aust., 

 is the same as Palceechinus, Seoul. MelecJiinus is Melonites, 

 Norw. & Owen. Cystocidaris, Zitt., is JEchinocystites, Wy. Th. 

 PalcBodiscus, Salter, which was intended for an Asterid genus, 

 really contains some Echinoidean types closely resembling Echi- 

 nocystites ; but the position of the periproct is not known. 



The genera remaining after these removals and absorptions 

 are classified under four orders of the subclass Palseechinoidea. 

 Bothriocidaris, a remarkable form from the Lower Silurian 

 formation, requires, as Zittel has shown, a special group, and an 

 order is established for it. The Perischoechinoida group them- 

 selves with as little friction as possible into two families, the 

 Archseocidaridse, with narrow ambulacra, and the Melonitidae, 

 with more than two vertical rows of poriferous plates in an 



