184 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



Phylogenetic facts must be acquired through the study of the comparative 

 anatomy of the group, combined with the study of the embryology and later 

 development; later they may be tested in tho light of the palaeontological record 

 if one so desires. 



In every group of animals we must be very careful how we apply the data 

 gained from the palseontological record, and in no group is this care more necessary 

 than in the crinoids. 



The earliest crinoids present many characters which are highly specialized, 

 and in general this specialization is along quite different lines from the specialization 

 in recent forms. Upon careful analysis they reduce themselves to a basic type 

 characterized by (1) a uniform cylindrical column of continuous growth; (2) a very 

 large calyx with an enlarged and asymmetrical anal area, including one or more 

 extra plates, and with a subradial plate beneath the right posterior radial; and 

 (3) short biserial arms. The post-palaeozoic crinoids, excepting the Encrinidae, upon 

 careful analysis reduce themselves to a basic type characterized by (1) a column 

 possessing a definite limit of growth and terminated proximally by a specialized 

 columnal with more or less of the character of a calyx plate; (2) a greatly reduced 

 and perfectly symmetrical calyx with no additional plates in the anal area and no 

 subradials; and (3) very long uniserial arms. In all three of these characters the 

 earlier crinoids are much more primitive than the later. 



The phylogenetic history of the crinoids, in agreement with the palaeontological 

 record and with the ontogeny, indicates that there has been a progressive and 

 rapid decrease in the size of the visceral mass, correlated with a proportionate 

 increase in the size and length of the arms. This reduction in the size of the vis- 

 ceral mass, and of the calyx plates, resulted in the eventual elimination from the 

 calyx of the subradials and of the interradials, leaving it composed only of the 

 infrabasals, basals and radials, while in the phylogenetically most advanced types 

 even the infrabasals, and in some extreme cases the basals also, have become meta- 

 morphosed or disappeared, so that the calyx is composed of radials only. 



We can not reconstruct the ancestral crinoid type from what we actually find 

 in the palaeozoic rocks, for every palaeozoic form is specialized in at least a minority 

 of its characters. For instance, in certain forms, in other ways possessing a com- 

 paratively high degree of specialization, the visceral mass has retained more or less 

 its primitive large size, so that we find the radianal (the right posterior subradial) 

 repeated under one or more, sometimes under all, of the other radials, as in For- 

 besiocrinus; while among the palaeozoic forms the majority possess a very primitive 

 type of column though there are several noteworthy exceptions, as for instance, 

 Platycrinus (fig. 516, pi. 1); many possess the primitive biserial type of arm, and a 

 few possess a very primitive type of calyx usually, however, combined with a 

 specialized type of arm. 



We must therefore reach our conclusions by a careful process of deduction, 

 and the result, arrived at through a critical study of the data presented by the 

 palaeozoic and later species, and especially by a study of the development and 

 morphology of the recent types, gives us an organism which, though closely ap- 



