THE SUDDEN ORIGIN OF NEW TYPES 415 



the food-grooves, by their closer structural relations to the 

 theca and in the more perfect development of basals and radials. 

 It will be noticed that in the early ancestral forms, such as 

 the Silurian Cystids Proteroblastus or Mesocystis, we meet 

 once more with a pronounced multiplicity of parts, in this 

 case consisting of an indefinite number of interradial plates 

 and the irregularly scattered diplopores. From these forms 

 the earlier Blastoids, such as the Ordovician Asteroblastus 

 and Blastoidocrinus (sometimes classed separately as Para- 

 blastoidea), seem to stand in close genetic relation ; among 

 other transitional characteristics their radials are not forked 

 and numerous small plates are intercalated between the radials 

 on the one hand and the large deltoids and ambulacra on 

 the other hand. The Blastoids, however, appear to have suf- 

 fered from a rapid over-specialisation and excess of regularity 

 and were consequently unable to survive the rigorous changes 

 of conditions at the close of the Carboniferous period, which 

 proved fatal to so many of the older types of animals. 



Pentamerism had also appeared already in certain Cystids, 

 e.g. Echinoencrinus among the Dichoporita, which are closely 

 related to the ancestral Crinoids ; and a distinct transitional 

 series to the regular pentamerism of Crinoids has been traced 

 in the Caryocrinidce^ from the Ordovician Hemicosmites to the 

 Silurian Caryocrinus, which exhibits close affinities to mono- 

 cyclic adunate Crinoids such as Hexacrinus of the Middle 

 Devonian. At the same time there is a concomitant re- 

 duction of the numerous brachioles to the usual five arms of 

 Crinoids. 



Pentamerism was no less well marked in Crinoids than in 

 Blastoids, but the plates of the calyx retained a more uniform 

 and generalised character than the highly specialised deltoids 

 and forked radials of Blastoids. In many cases, however, a still 

 further reduction in number took place in Crinoids, especially 

 in the dicyclic forms, in which the infrabasals may be reduced 

 to four, or more usually to three, upon which the five basals 

 are superposed. Even the basals in the dicyclic forms become 

 also subject to a reduction to three, evidently for the purpose 

 of giving greater strength to the calyx ; but the radials of 

 course always remain five in number, corresponding to the 

 invariable and constant number of the arms. Infrabasals no 

 longer occur so frequently nor are they so well developed 



