248 



ECHINODERMATA— ASTEROZOA 



PHYLUM IV 



Starfishes, are still uncertain. All other Starfishes may be grouped, as was 

 done by Sladen, in a contrasting order, Cryptozonia ; but this is probably not 

 a natural group, and Fisher distinguishes two divisions. These, however, are 

 separated by characters not ascertainable in Paleozoic remains, and for practical 

 purposes, the paleontologist may well accept the Cryptozonia. 



Order 1. PHANBROZONIA. 



Asteroidea in which the marginal ])lates are large and conspicuous ; papulae 

 nearly always confined to the dorsal surface; ambulacral ossicles not crowded, and 

 tube feet in two rows in each ambulacral groove. 



This order includes a large proportion of the Paleozoic and Mesozoic 

 Starfishes, besides numerous Kecent genera. Fisher groups the Eecent forms 

 in no less than a dozen families, and Gregory gives four others for Paleozoic 

 species alone. Many of the Recent families are not known as fossils, and 

 others may have one or a few extinct representatives. The most important 

 families and genera from a paleontological point of view are the following. 



Family 1. Palaeasteridae. 



The typical members of this family have the ambulacral ossicles more or 

 less completely alternating. The adanibulacral plates are most conspicuous 



in the mouth parts. The 

 marginal plates and many 

 of the abactinal plates are 

 conspicuous. 



The exact limits of this 

 family are hard to determine, 

 as the known forms are all 

 from the lower Paleozoic. 

 The typical genus Falaeaster 

 Hall (Fig. 351) is known 

 from both Europe and 

 America. Hudson has re- 

 cently proposed a new order 

 and family for an interesting 

 Starfish (Protopalaeaster) 

 from the Ordovician rocks of 

 Canada. It is, however, 

 probably allied to Palaeaster. 

 The i^lates which Hudson 

 calls epineurals are probably 

 the amhulacrals seen from within, the dorsal side of the animal being lost. 



Several genera allied to Palaeaster have been described (Petraster Billings ; 

 Archasterias Midler ; Argaster Hall ; Ataxaster Jaekel, etc.), while other less 

 typical forms (Xenaster Simonowitsch ; Tetraster Eth. and Nich.) have been the 

 subject of debate as to their true position. Whether Lindsfroemaster Gregory 

 is properly referable to this family is not certain ; it bears considerable super- 

 ficial resemblance to the Goniasteridae. 



rn-.-ri.v 



bhi. 351. 



Palaeaster enckarls Hall. Devonian; Hamilton, New York. A, 

 Ventral aspect, natural size. B, Dorsal surface of one of the arms. 

 (•', Diagrannnatic view of ventral surface of the arms (after Hall"). 



