ZOOLOGICAL CI1AKACTKRS. lli 1 



with larger deltoids which meet the radials along a horizontal suture al the level of 

 the distal ends of the amhulacra. According to Schmidt the lower edge of the 



deltoid is marked by hydrospire-slits, which do Dot, however, extend down on i<> the 

 upper edges of the radials beneath. The summit would thus resemble thai of 

 Codaster trilobatus if it were raised into a low dome and had triangular instead of 



rhombic deltoid plates (PI. XIII. 6gS. I, 3, 4). "We should thus have no difficultj in 



placing Blastoidocrinus among the Blastoidea, though it would have to be assigned 

 to an altogether separate family from the other members of the group, on account of 



the singular position of the radio-deltoid suture and the restriction of its hydrospires 

 to the deltoid [dates; but in neither of these characters would it offer an easy transi- 

 tion from the Blastoids to either Crinoids or Cystids. We must confess to some 

 doubt, however, respecting the presence of its hydrospire-slits as described 1>\ Schmidt, 

 for we have examined a deltoid plate of a Canadian specimen which appears to presenl 

 the same structure; and this, as revealed by a section, is certainly nothing like the 

 hydrospire-apparatus of a Blastoid. We are much rather inclined to think that it is 

 a surface ornamentation, more especially as we have noticed a bifurcation of some of 

 the ridges separating the supposed slits; and we know of no hydrospires, either in 

 Blastoids or Cystids, in which such a structure occurs. 



At the same time we are quite prepared, upon production of sufficient evidence, to 

 accord Blastoidocrinus a place among the Blastoidea; and in some respects it would 

 agree much more closely with the typical Blastoids than either Astrocrinus or 

 F.h utherocrinus. 



Regarding the group Blastoidea as a class of Echinoderms, we have now to consider 

 how it may further be subdivided into orders and families. An attempt in this 

 direction has already been made by S. A. Miller 1 , who gives the group only ordinal 

 value and divides it into three families, neither of which, however, does he define ; 

 and his names therefore have no real systematic value. His first family, Nucleocrinida?, 

 contains the single genus Elaacrinus or, as he prefers to call it, Nucleocrinus. In the 

 second family, Pentremitidae, are included the genera Blastoidocrinus, Granatocrinus, 

 Peutremitcs, and Troostocrinus ; while the third family, Stephanocrinidpe, embraces 

 the four genera Codaster, Codonites (which is more correctly known as OrophQcrinus), 

 Eleutherocrinus, and Stephanocriwus. Three of these genera, however, represent 

 widely different morphological types; and we cannot help thinking that if Mr. Miller 

 had attempted to define the families which he established so easily, he would have had 

 some difficulty in finding any characters common to Kleutlicrocrinus (PI. XIX. figs. 4-G) 

 and to Codaster (PI. XII. figs. 1-6; PI. XIII. figs. 1-3) other than those which are 

 possessed by all Blastoidea. 



It is not difficult to divide the class Blastoidea into two orders ; for JSleutherocrinus, 

 Astrocrinus, and Pentephyllum are very strikingly different from all the other types 

 of Blastoids which were known to Roemer. These last have five equal and similar 

 1 American Palaeozoic Fossils, Catalogue of the Genera and Species, 2nd edition, 1SJSI5, p. 277. 



B 



