72 THE CRINOIDEA CAMERATA OF NORTH AMERICA. 



Anomalocrinus the left iDOsterior radial, which is the largest plate in the calyx, 

 is sometimes bisected longitudinally, but this is not constant. 



The gradual increase in size of the upper segments, and the disappearance 

 of compound plates in other rays than the p/Dsterior, indicates that in these 

 Crinoids there is a transition from three compound plates to one, and also 

 from compound to simple, plates ; and we think it may be safely assumed 

 that there was a time in the early history of the Crinoids when the arm- 

 bearing section was altogether unrepresented. This was apparently the case 

 in Baerocrinus, in which two of the radial plates are non-arm-bearing, and as 

 these plates occur in the same rays as the compound plates of Anomalocrinus, 

 we may infer that Baerocrinus is the ancestral form, lower in its development 

 than either Anomalocrinus, Hoplocrinus or locrinus. 



The later Fistulata have no true compound radials; and although the 

 lower segment of the posterior radial is still represented in many of them, that 

 plate does not bear the same relation to the arm-bearing plate that it did in 

 the earlier forms, but assumes the functions of an anal plate. As such, it will 

 be considered by us in its proper place. 



Among the Larviformia, true compound radials occur in Hajylocrimis, which 

 has three of them : in the right posterior, and the right and left antero-lateral 

 rays, — the same rays in which they occur in the Heterocrinidse. The com- 

 pound plates of the Ichthyocrinidae will be discussed in connection with, the 

 anal plates. 



In groups in which the radials are simple, the five plates are approxi- 

 mately of the same size ; but when compound, they often differ considerably 

 in size and form. This is most apparent in Calceocriniis, Catillocriniis, Haplocri- 

 mis, and Pisocrinus. Among the Poteriocrinidge also, when the inferradial is 

 represented, the right posterior one is smaller than the others, and of some- 

 what different form. In some Palaeocrinoidea, the radials are separated by 

 supplementary plates ; while in all later Crinoids, with the exception of the 

 recent genus Thaiimatocrimis, they are in contact all round. In the Ehodocri- 

 nidge, and curiously enough in Thaumatocrinus, they alternate with large inter- 

 radial plates, with which they form a ring often pieces. In the Thysanocrinidoe, 

 Actinocrinidae, Batocrinidge, Acrocrinidae, Dichocrinid^, and in most of the 

 Fistulata and Ichthyocrinidse, a special anal plate is interposed between the 

 two posterior radials. In the Eeteocrinidae, they are separated by numerous 

 small, ill-defined plates at the four regular sides, with the addition of an anal 

 plate at the posterior side. In the Melocrinid^, Eucalyptocrinidae, Platycri- 



