MONOGRAPH OF THE EXISTING CRINOIDS. 



107 



doubting that the first brachial pair, normally situated at the base of the free 

 undivided arm, is here represented by the third and fourth postaxillary ossicles, 

 while the first and second represent a reduplication of this pair situated between 

 it and the axillarj' or distal element of the most proximal pair. 



The structure of the arm of Ulntacrhius from the axillary onward, therefore, 

 is identical with that of the entire arm in Eudiocrinus. A reduplication of the 

 first brachial pair is present in which the distal element is not axillary. 



Judging from its universal presence in all other genera, the absence of the 

 second arm on the reduplication of the first brachial pair situated between the 

 normal first brachial pair and the axillary would seem to be a retrogressive 

 character; we should therefore expect that in the young of Vintacrinus the second 

 postaxillary ossicle would appear more like an axillary than in the adult, and 

 we find that such is the case. 



The correctness of this interpretation of the arm structure of Vintacrinus is 

 shown by a specimen with 11 arms in which one of the second postaxillary 

 ossicles is an axillary bearing two arms instead of an arm and a pinnule as 

 usual; and another, with nine arms, in which the second postradial ossicle 

 instead of being axillary bears an arm and a pinnule, so that three similar pairs 

 of ossicles, each bearing a continuation of the postradial series and a pinnule 

 on the distal border of the outer element, are present. 



Of interest in this connection is a specimen of Florometra perplexa (fig. 131, 

 p. 79) at hand, in which on one of the arms arising fi-om a IBr axiUary and 

 immediately following the latter there are two pairs of ossicles exactly alike. 

 This arm therefore possesses the structure characteristic of the arms of Vintacrinus. 



Fentoiuetrocrinidw and Atopocrintis. 



In ThaumatocHnus (part 1, figs. 113, 114, p. 181; 115-118, p. 183), Penta- 

 mrtrocrinm (fig. 126, p. 79, and part 1, figs. 119, p. 185; 120, p. 187; and 121, 

 p. 189), and Atopociinus (part 1, fig. 227, p. 245) the arms are undivided. Exami- 

 nation shows that the first two ossicles following the radials correspond exactly 

 to the first two brachials in the free undivided arms of other types. 



Therefore Thaumatocrlnus^ Pentavietrocrinus, and Atopoc7'inus are simple 

 forms in which the IBr series are absent, and the free undivided arms arise 

 directly from the radials. 



The species of these three genera have the simplest arm structure of any of 

 the comatulids. 



In Thaumatocrinus and Pentametrocrin'us, and to a lesser degree in Atopo- 

 ciinus, the ventral disk runs for a considerable distance up the arms, these radial 

 tapering prolongations giving it a sharply stellate shape, and merges into the 

 ventral perisome of the arms at a point far beyond the second brachial. In no 

 other types does this occur. Also in these three genera, alone among the recent 

 comatulids, the first syzygial pair does not immediately follow the first two 

 Iirachials but is separated from them by an interpolated brachial. 



