lO SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 76 



which we know to be invariably present, being in some cases by no 

 means distinct. 



Furthermore, there is undoubtedly in this family a strong tendency 

 to fusion of basals, which causes considerable uncertainty in the 

 composition of the base as we find it in the fossil state. This is 

 nothing unusual. In the extensive family Platycrinidae, which we 

 know beyond question has primarily a tripartite base, there are many 

 species in which the plates are completely fused into an undivided 

 disk. Wanner finds both the divided and undivided base among the 

 Timor species of Platycrinus, and of his new genera Eutelecrinus and 

 Neoplatycrinus. In the Devonian Triacrinus, in which the three 

 basals was the sole generic character specified to separate it from 

 Pisocrinus, my specimens from the Eifel show 3, 4, and 5 BB. An 

 excellent example showing how this character appears in practice is 

 furnished by a species of Synbathocriniis, S. robustus, occurring in 

 the same beds and localities, and under identical conditions of preser- 

 vation, with Catillocrinus tennesseeae ; out of 55 specimens, 16 show 

 distinctly all three sutures of the divided base, six show one or two 

 faintly, while in the remaining 33 the base appears undivided. 



In the Burlington species of Catillocrinus, C. wachsmuthi, in which 

 the basal ring externally appears much the same as in C. tennesseeae, 

 though usually somewhat less concealed by the column, there is 

 ample evidence of a divided base, although all the sutures are not 

 usually seen, In three specimens there is a complete division into 

 3 BB, the sutures meeting at the axial lumen, and leaving no possible 

 space (at the exterior) for infrabasals (pi. 3, figs. 7, 10). So also in 

 the later species of the Keokuk and Chester groups, in which the 

 basal ring is relatively higher and laterally more visible, three well- 

 defined sutures are seen in several specimens, while in others of the 

 same species none can be detected. 



Therefore I think it may fairly be concluded from the evidence 

 that the basal ring in some species of this genus is divided into at 

 least three plates, but that owing to the tendency to fusion a partial 

 or complete anchylosis may at any time occur, leaving the observer 

 in doubt as to the actual number of plates. This fact must be strongly 

 emphasized, because the negative evidence as to the presence of 

 sutures in these forms is liable to be extremely misleading, as I have 

 learned from some practical tests. In two fragmentary specimens of 

 C. tennesseeae, I carefully prepared the inner surface of parts of the 

 calyx opposite to some of the most prominent sutures, such as those 

 between the radials, always very conspicuous at the exterior. They 



