80 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.41. 



to the theca. In the present case there has probably been a dis- 

 ruption of the column in its proximal portion succeeded by a partial 

 resorption of the columnals. The result of this process is the forma- 

 tion of a structure closely approximating to a centro-dorsal. The 

 specimen of P. sorlinensis, figured by de Loriol in the same volume 

 (PL 205), apparently may be explained as having the same type 

 of structure. In this case the columnals bear numerous long cirri 

 which had probably not yet been acquired in the specimen previously 

 described. Such a disruption of the column in the case of Penta- 

 crinus, taken in connection with what we know relative to the wide 

 maintenance of a detached existence by these forms, indicates that 

 there was a strong tendency among the Pentacrinidse toward the 

 formation of essentially comatulid types. Such forms as those 

 described above need but a fusion of the columnals, and an increase 

 in the number and specialization of the cirri to make very fair 

 Comatulie. 



Another feature which is not of great importance but nevertheless 

 is somewhat suggestive of a similar derivation is the distinctly 

 pentagonal section of many of the fossil comatulid centro-dorsals. 

 This, for instance, may be noted in Antedon (Solanocrinus) scrohicu- 

 lata (PL 8, fig. 3), in Antedon dejJressa (PL 8, fig. 8), and in many 

 other fossil comatulids as figured by various authors. In the case 

 of A. scrobiculata it is further to be noted that the cirrus sockets have 

 an alternating "biserial" arrangement which points rather strongly 

 to the conclusion that each successive cirrus on any given face of the 

 centro-dorsal pertains to a distinct columnal. The pentagonal cross- 

 section of such centro-dorsals may be held as resulting from secondary 

 changes as induced by the presence of cirri, but comparison of these 

 centro-dorsals with subglobose or bowl-shaped types seems to indicate 

 that this may not necessarily be the case. Indeed the tendency 

 toward modification of the shape of the centro-dorsal points rather 

 toward the acquisition of a quite dissimilar type of organ. 



It is not impossible that we may look among the species of Milleri- 

 crinus, as defined by de Loriol, for ancestors of certain comatulid 

 lines. This genus, as defined by de Loriol, contains many highly 

 divergent types which at best may not be held as congeneric. Certain 

 species tend toward Apiocrinus, which was doubtless derived from 

 these forms, while others simulate in a marked degree members of the 

 Pentacrinidse. The genus is not as far removed from the Penta- 

 crinidse as is currently held, but whether these strikingly similar 

 types represent members of quite distinct phyletic lines or are the 

 result of convergence it is impossible to say. The evidence certainly 

 points to a common ancestor not far removed. Taking Millericrinus 

 in its accepted sense one might explain the petaloid markings of 

 certain Comatula?, as noted above, as the result of derivation from 



