NO. 1846. ON CERTAIN ELEVTHEROZOIC PELMATOZOA—KIRE. 79 



when one considers the Bourgueticrinus type of colums possessed, and 

 the fact that the stem lumen is circular in section (PI. 7, fig. 12). * 



The evidence relative to the nature of the perforation of the cen- 

 tro-dorsal being inconclusive because of our lack of exact knowledge, 

 it is necessary to fall back on other more obvious and satisfactory 

 characters. On Plate 8, figure 7, is given a view of the lower surface 

 of the theca of Antedon depressa from which the centro-dorsal has 

 been removed. Figure 8 shows the normal appearance of a cup with 

 the centro-dorsal attached. To be compared with this is the basal 

 view of the theca of Pentacrinus (Extracrinus) dargniesi (PI. 8, 

 fig. 5). In both forms we may note a general similarity in structure, 

 particularly as regards the development of the basals. The basals of 

 Antedon in this instance, however, quite lack the petaloid markings 

 characteristic of the Pentacrinus and the Pentacrinidse in general. 

 The specimen of Antedon (Solanocrinus) scrohiculata here figured 

 (PI. 8, fig. 4) is noteworthy in this regard. As may be seen, the 

 articular faces of the basals have the characteristic markings of the 

 same plates in the case of the Pentacrinidse. Carpenter (18806) has 

 figured Comatulse exhibiting the same type of structure. Such for 

 instance is his Adinometra cJieltonensis (PI. 11, fig. 206). He like- 

 wise figures a specimen of Antedon (Solanocrinus) scrohiculata. in 

 which the markings are very distinct. This type of articulation 

 may only I' think be attributed to forms in which a pentalobate 

 column is possessed at that point. The natural inference is that 

 the stalked ancestors of such types are referable to the Pentacrinidae, 

 although this need not necessarily follow. It seems impossible, how- 

 ever, to derive these types directly from the Bourgueticrinidae. 



In connection with the statement made above relative to the possi- 

 bihty of Pentacrinus or some member of the Pentacrinidae having 

 formed the radical from which might have sprung a line of coma- 

 tulids, should be noted certain curious structural features in the case 

 of Pentacrinus (Extracrinus) collenoti as figured by de Loriol (1884-89, 

 PI. 199, figs. 3, 3a, 36) showing a most remarkable modification of 

 the normal Pentacrinus stem structure. In this form it is to be noted 

 that the column in its proximal portion is of normal size, but distad 

 it contracts rapidly forming a stump of exceeding brevity. This 

 stem fragment consists of something like 10 columnals. In every 

 way it reminds one of conditions as they are to be found in Milleri- 

 crinus prattii where a few of the proximal columnals remain attached 



I The tjrpe of coliimn possessed by the larval comatulid is after all directly comparable to that to be found 

 In the very young Isocrimis. As figured by Clark (1908, p. 88, text figs. 3 and 4), the distal portion of the 

 column of Isocrinus dccorus has oval columnals, essentially similar to those here figured in the case of Ante- 

 don. This fact again indicates a close relationship of the Pentacrinidae and at least certain lines of the 

 Comatulse. Indeed in both types there is an essential unity of structure up to the time detacliment from 

 the larval column becomes effective. Subsequent to that time the critical difference is that the Penta- 

 crinidae form new columnals and maintain a column whereas the comatulids do not. 



