MORPHOLOGY 20. 



desirable ; the characters are not, in practice, of much service .in the differen- 

 tiation of groups, the stem in this order, as in the Paleozoic crinoids generally, 

 not having attained that fixity of structure which renders it an element of such 

 great importance in the Recent crinoids. Occasionally, however, a stem char- 

 acter is remarkably constant for a species; for example in Taxocrimts colletti 

 from Crawfordsville, Indiana, the most abundant species of a Flexible crinoid 

 known, where the enlargement next to the calyx, instead of tapering gradually 

 downward until it merges into the cylindrical stem, is invariably short and 

 thick, ending abruptly with a rounded offset (Pis. LVI and LVII). 



There is no reason for thinking that the mode of growth of the stem in the 

 Flexibilia is essentially different from that in the other permanently stalked 

 crinoids, viz., by means of new ossicles introduced next to the calyx, or inter- 

 polated between older ossicles within a short distance down. Upon this point 

 the view of Wachsmuth and Springer expressed in the North American 

 Crinoidea Camerata, pp. 39, 40, viz., that in this group the topmost columnal 

 was persistently fused with the infrabasals so that the column increased by the 

 interpolation of new ossicles below it, is in my opinion more theoretical than 

 practical. That such a fusion often occurred in adult specimens in which stem 

 growth was completed is doubtless true, as is indicated by specimens like those 

 of figure 2a, Plate XXVI, and figure 9, Plate LXXIV. But those occurrences, 

 where the infrabasals became detached from the rest of the calyx and adhered 

 to the stem, are rather to be explained by the nature of the union between the 

 plates: (1) that between infrabasals and the succeeding basals being by loose 

 suture, with large and deep ligament fossae and small contact areas ( PL XXIV, 

 figs, la, b) ; while (2) that between infrabasals and stem was by plane sur- 

 faces with very fine radiating striae, i. e., syzygy — a closer union than that of 

 the stem ossicles with one another. The second connection was therefore 

 stronger than the first, and with decomposition the infrabasals would separate 

 from the basals rather than from the stem. The proximal ossicles in most of 

 the Flexibilia were exceedingly thin, sometimes translucent in the fossil 

 (PI. XXX, fig. 5a), and their often persistent attachment to the infrabasals can 

 readily be accounted for by a close suture without need of invoking an actual 

 fusion, although this also may have occurred in mature individuals. 



Such attachments are frequently found in Forbesiocrinus and Onycho- 

 crinus; but in many other genera nothing of the kind occurred, the stem sepa- 

 rating from the calyx by a perfectly smooth surface (not of fracture). This 

 was probably the case in most of the Silurian genera, in which the stem in- 

 creased by alternating long and short columnals without proximal enlargement ; 

 and it is notably so in the Carboniferous genus Ampliicrinus, in which every 

 one of over 20 specimens has the topmost stem ossicle clearly separated from 

 the calyx. It was probably the case also in the' entire family Ichthyocrinidae, 



