264 CATALOGUE OF THE BLASTOIDEA. 



fig. 1). Thus then we believe that the principal character on which Hcteroschisma was 

 established has no existence in fact. The only point in which there is any difference 

 between Ileteroschisma, as typified by H. gracile, Wachsmuth, and an ordinary Codaster 

 is altogether of minor importance. Mr. Wachsmuth says that the side plates of the 

 former meet the sides of the sinuses, though only " rarely " so in //. gracile 1 . 



We think that this may perhaps be the case, though we are by no means convinced 

 that it is so. It is possibly due to the greater depth of the radial sinuses and their 

 consequent diminution in width. We have noted what appear to be precisely similar 

 differences amongst the various species of Phcenoschisma. Under these circumstances 

 we feel with great regret that we cannot accept Ileteroschisma as a valid genus, but 

 must simply look upon it as a synonym of Codaster. 



Codaster was established by McCoy as a Blastoid, but Mr. Rofe 2 regarded it as a 

 connecting link between the Crinoidea and Cystidea, Pentremites being more closely 

 allied to the former, and Codaster to the latter. On the other hand, Billings 3 

 definitely referred it to the Cystidea, because there is no connection between its 

 hydrospires and the cavities of the pinnulse borne on the ambulacra, such as he 

 assumed to exist in Pentremites. We cannot learn that any other palaeontologist 

 but Prof. K. Zittel 4 has definitely adopted this view, in which we certainly do not 

 agree. If Codaster be a Cystid, so are Orophocrinus and Phcenoschisma, which are 

 also devoid of ambulacral pores. But Codaster is a true Blastoid in every respect ; 

 and we think that Billings was led to this erroneous conception by his not having 

 emancipated himself from the old doctrine that the cavities of the pinnulse and 

 hydrospires of a Blastoid are directly connected by the ambulacral pores. Neither 

 Codaster, Orophocrinus, nor Phoenoschisma has any such pores, but the relations 

 between them and the remaining Blastoids are so close that there can be no question 

 of their systematic position, while Billings's theory is now abandoned by all the more 

 prominent writers on the Blastoidea. 



Species. After mature consideration we are prepared to admit five, or perhaps nine, 

 species of Codaster, though not more, on account of the indefinite manner in which 

 certain so-called Codasters have been described and figured. In Britain we possess 

 only one, C. trilobatus, McCoy, but it presents several varieties (PI. XIII. ). The 

 American species differ from the British type in possessing a more elongate form, 

 greater convexity of summit, a narrower base, and more complex ambulacra (PI. XII. 

 figs. 1-6). Four of these are well defined, viz. C. altematus, C. gracile, C. Hindei, 

 and C. pyramidatus. Shumard described another species together with the last 

 named as C. Americanus, but said that it was " very similar to C. pyramidatus, of 



1 Keport Geol. Survey Illinois, 1883, vol. vii. p. 356. 



2 Geol. Hag. 1805, vol. ii. p. 251. 



3 American Journ. Sci. 1809, vol. xlviii. p. 80. 

 * Handb. Pal. 1879, 1 Bd. Lief. 3, p. 424. 



