122 CATALOGUE OF THE BLASTOIDEA. 



radials and five ambulacra of the same character, which all meet at the peristome 

 at equal angles (PL III. fig. 14; PL IV. fig. 17 ; PL V. fig. 23 ; PL VII. figs. 5, 14; 

 PL IX. figs. 14, 15; PI. XIII. figs. 1, 4; PL XV. figs. 2, 11). In the Astro- 

 crinidse, however, there is not only an abnormal development of the basals (PI. XIX. 

 fig. 5 ; PL XX. figs. 1, S, 13, 15), but one radial and the corresponding ambulacrum 

 are altogether different from their fellows (PL XIX. figs. 1, 4-6; PL XX. 

 figs. 0, 13-20). The angle at which this azygos ambulacrum joins the peristome 

 is different from the other four interambulacral angles, and its structure is altogether 

 abnormal. The loss of a perfect radial symmetry which accompanies these pecu- 

 liarities has led us to separate the typical Astrocrinidse from the more regular 

 Blastoids as an Order " Irregularis "; and with them we associate the problematical 

 Pentephyllum, which closely resembles Eleutherocrinus in the distribution of its 

 ambulacra (PL XVI. fig. 14 ; PL XIX. fig. 6), though it seems to have a tolerably 

 symmetrical base (PL XVI. fig. 16). The basiradial sutures are not so clear as we 

 could wish, but so far as we can judge from the only specimen known, which is 

 merely an internal cast, there does not appear to be so much difference between the 

 azygos radial and its fellows as in the typical Astrocrinidge. 



Taking then the Blastoids known to Roemer as constituting the Order " Regulares," 

 we have now to see how they can be grouped into families. Roemer 1 proposed 

 to make four different groups of the genus Pentremites, which he called respectively 

 Floreales, Elliptici, Truncati, and Clavati. The type species in each of these 

 groups is now generally regarded as representing a distinct genus, viz. Pentremites, 

 Granatocrinus, Pentremitidea, and Troostocrinus ; and this shows better than any- 

 thing else could the truly natural character of Roemer's classification. 



Pentremites and Pentremitidea are regarded by us as constituting, together with 

 our new genus Mesoblastus, the family Pentremitidae ; while Troostocrinus and 

 Granatocrinus are the types of the Troostoblastidai and Granatoblastidai respectively. 

 Codaster and Elceacrinits, which were considered by Roemer to be altogether distinct 

 genera, also appear to us to represent morphological types of family value ; and in 

 each case there are other more recently established genera which may fitly be 

 placed in the same family. 



All the six types of Blastoids, therefore, which were recognized by Roemer 

 appear to us to be of generic value, while five of them are the leading genera of 

 families. 



Since the differences between his four groups of the genus Pentremites are princi- 

 pally founded "nach der ausseren Gestalt des Kelches und der Pseudambulacral- 

 Felder im Besonderen," it is only to be expected that they should eacli contain 

 certain species which the progress of knowledge lias shown to belong to an altogether 



1 Archiv f. Naturgesch. Jahrg. xvii. 1851, lid. i. p. 352. 



