REPORT ON THE CRINOIDEA. 23 



process and represent the unabsorbed lateral portions of the primary layer forming 

 the embryonic basal plate. The ventral side of the basal ray in the three species 

 of Actinometra which are figured on PL IV. figs. 5a, 46, Gb, a, is marked by a 

 relatively large depression which forms the central end of the axial interradial canal. 

 This descends into the calyx over the apposed lateral edges of two radials, as is well 

 seen in PI. III. figs. Id, 7c, and PL V. fig. 2c. But in most cases it ends blindly 

 without reaching the dorsal surface of the radial pentagon at alL 



D. The Radials. 



The radials of Comatulse differ considerably from those of the Pentacrinidae, the 

 family of Stalked Crinoids to which the free forms are most closely allied. In Pentacrinus, 

 as in the Pentacrinoid larva of Antedon (PL XIV. figs. 2-9), the first radials appear 

 above the basals on the exterior of the calyx as relatively large convex plates. They 

 retain this character in Thaamatocrinus and to a less degree in Atelecrinus (PL VI. figs. 

 5, 7). both of which are permanent larval forms in other respects. But in the two 

 large genera Antedon and Actinometra there is a very considerable amount of variation 

 in the extent to which the first radials appear externally. Some forms, such as Antedon 

 eschrichti, show no indication of them at all (PL XXIV. fig. 11), or only traces of their 

 angles in the interradial portions of the calyx (PL I. fig. 8a) ; while in other cases, such 

 as Antedon elegans, Antedon longicirra (Pis. VIII. , XVII.), and Antedon macronema 

 (PL IV. fig. 3a; PL XXXVIII. fig. 5), they exhibit a relatively large outer surface between 

 the edge of the centro-dorsal and the second radials. Between these two extremes every 

 intermediate gradation may be traced. The former is due to the gradual enlargement of 

 the centro-dorsal, which spreads itself over the base of the calyx towards the end of 

 Pentacrinoid life and sometimes conceals the first radials altogether, as described by Dr. 

 Carpenter 1 in Antedon rosacea. The second radials thus appear to spring directly from 

 the centro-dorsal (PL XIII. fig. 2); and this has sometimes led to species of Comatidse 

 being described as having only two radials. In fact d'Orbigny 2 described a new genus 

 Comatulina, for a fossil species in which there are no external basals and " les bras 

 s'articulent immediatement sans intermediates a la piece centrale pourvue de ramules." 

 The full-grown Antedon accela presents this appearance, but the younger specimen 

 figured on the same plate (PL XVI.) shows comparatively large first radials; while a 

 more mature individual (PL II. figs. 3a, 3c) shows relatively less of them, owing to the 

 spread of the centro-dorsal. The figures of Antedon plialangivm on PL XXVIII. 

 show similar differences of growth, though in a less degree. 



The various species of Actinometra exhibit among themselves essentially the same 



1 Phil. Trans., 1866, p. 742. 



2 Cours elementaire de Paleontologie et de Geologie Stratigraphique, 1850-52, vol. ii. p. 139. 



