262 SMITHSONIAN INSTITUTION 



The rudimentary and more or less unstable condition of the infrabasals is 

 the chief diagnostic character of this family. These plates, which in the 

 Lecanocrinidae form an essential part of the calyx, are here so completely with- 

 drawn from the exterior that in specimens with the column attached there is no 

 evidence of their existence. In typical genera they form a mere vertical plug, 1 

 of no functional importance; and in several species this plug is completely 

 resorbed, leaving an open space of substantially its own form and size. Cor- 

 related with the foregoing is another character equally constant, which gives 

 to the genera of this family a certain conspicuous facies by which they can 

 usually be recognized from any fragment containing the lower portion of a ray ; 

 this is the relatively small size of the radial plates. With this is combined a 

 close abutting or interlocking of the rays either from their origin or above the 

 interbrachials when present, producing a continuous wall for several series of 

 brachials. Consequently the rays are fan-shaped, widening from the radials 

 up. The presence of interbrachials does not materially alter this peculiar 

 aspect. There is, therefore, in this family in its most extreme form that non- 

 differentiation of the calyx which to a greater or less degree characterizes the 

 order Flexibilia, the arms being mere continuations of the lower brachial series, 

 which pass into them without any apparent change except gradual reduction 

 in size. 



By reason of the correlation of these characters the Ichthyocrinidae form 

 a remarkably compact, well-defined family group, which is sharply distinguished 

 from the two preceding, although it includes both rotund and elongate forms. 

 Only one of the genera, Clidochirns, might be confounded with the Lecano- 

 crinidae, owing to variability in size of the infrabasals among the species ; while 

 Synerocrinus, although thoroughly typical as to this character, on account of 

 its weak anal side is a transition form toward the Taxocrinidae. 



There is in this family a very notable suppression of anal structures, the 

 pentamerous symmetry of the calyx being but little modified by them. In four 

 of the genera the posterior basal is not differentiated, and in a fifth, Synapto- 

 crinuSj although elongated it is not followed by any anal plate. In one, Amphi- 

 crinus, the posterior interradius is smaller than the other four. The radianal is 

 found in only three genera (Silurian and Devonian) in the strictly primitive 

 position as an inferradial, and is so inconspicuous that its existence as a calyx 

 element was never recognized until I pointed it out in 1906; it is most promi- 

 nent in some species of the otherwise aberrant Clidochirus. The oblique form 

 of radianal is not found at all. Two Silurian, two Devonian, and all the Car- 

 boniferous genera are without any identifiable radianal. 



The dichotomous arm structure prevails in a large majority of the genera, 

 but in the three in which the heterotomous type occurs it attains the most per- 

 fect stage with conspicuous main rami fringed with subordinate ramules. 



1 Text-fig. 11, p. 117. 



