154 THE CRINOIDEA CAMERATA OF NORTH AMERICA. 



cumstances, and about all we can say of the " value " of any character is 

 that its value is proportional to the extent to which it tends to combine 

 subordinate groups. As a rule we may say that a character which passes 

 down to earlier geological epochs, is of greater value in classification than 

 one dating from a later period. 



It has been stated that the supplementary plates afford excellent charac- 

 ters for dividing the Crinoids into primary divisions ; but they are also of 

 great importance in the separation of families. It may appear singular that 

 plates whose chief function it is to fill up spaces between other plates, and 

 which are of but secondary importance morphologically, should have so 

 great a value in classification ; but such is undoubtedly the case. The elimi- 

 nation of those plates from the calyx would reduce all Crinoids to a single 

 group, and these would all be Larviformia. The introduction of supple- 

 mentary plates into the simplest Inadunate type produces the Fistulata, 

 Camerata, and Articulata, and all in their individual development necessarily 

 passed through the Larviformia stage. 



Among the supplementary plates, the anals unquestionably take the first 

 rank. The absence of anal plates, their introduction into the calyx, and the 

 relations they bear to adjoining plates — w^hether introduced between the 

 radials or brachials, or between both of them — has been regarded as very 

 important ; but the same characters have not the same classificatory value 

 in every group. The differentiations produced by the anal plates have been 

 considered of family importance among the Camerata, but among the Fistu- 

 lata and Impinnata appear to be of generic value only. 



The nature of the base, whether composed of one ring of plates or two, is 

 a valuable character for distinguishing families, but we thought it expedient 

 to make an exception to this rule in the case of the Eeteocrinidse, under 

 which we have united monocyclic and dicyclic forms. The number of basals 

 and infrabasals is of generic value only. 



The occurrence of inferradials among Inadunata and Articulata, and the 

 modifications they undergo palasontologically, and their final disappearance, 

 have furnished good generic characters. 



Somewhat less important from a classificatory standpoint is the number 

 of costals, although it was made a generic character among the Camerata ; 

 while in the Fistulata it has in some cases no significance at all. 



The general structure of the disk affords good characters for dividing the 

 Crinoids into ordinal and subordinal groups. It was the total absence of 



