23O SMITHSONIAN INSTITUTION 



above the line indicated, and the striking curvature of the distal margin of the interbrachial 

 areas, together with the total absence of angular faces for interlocking plates there or at 

 the margins of the rays above, all negatived this supposition. 



A careful study of the markings on many plates showed that this peculiar rugose margin 

 could not be a sutural surface in the usual sense. It is not in a flat plane as we must expect 

 in such parts of sutural surfaces as are apposed ; there is a slight curvature in a dorso-ventral 

 direction, and also a gentle rounding of the margins longitudinally, which are in sharp con- 

 trast to the condition of the interlocking sutures in Forbesiocrinus agassizi and F. saffordi 

 (PI. XXI, figs. 2, 3). There is a slight, but perceptible, projecting rim at the dorsal edge 

 of the plates (PI. XXIII, fig. id), and a careful examination of the grooves forming the 

 corrugated surfaces shows that they are rounded, with flowing lines by which in the upper 

 part they connect with, or are merged into, the peculiar longitudinal grooves at the ventral 

 side of the ray (PL XXIV, figs. 17b, 22). These markings may be profitably compared 

 with those on the margin of the radials in Pycnosaccus (PL XI, fig. 9c) which are definitely 

 known to form the surface for attachment of interbrachial perisome. 



It is evident, therefore, that they represent lines of growth of the calcified integument 

 forming the perisome of the upper part of the interbrachial areas, which in this case begins 

 on the outer or dorsal edge of the plates. This produces a structure apparently different 

 from that found in other genera in the thickness of the edge connecting with the perisome, 

 no doubt due to the greater thickness of the perisome itself. We may have a clew to its 

 origin in Lithocrinus, which has a similar large first interbrachial followed by other ranges, 

 the plates of which begin to be irregularly isolated by perisome surrounding them. The 

 perisome is on a level with their outer surface, and the imbedded plates are presumably as 

 thick as those which are suturally united ; therefore if we could examine the sides of one of 

 them we should expect to find a corrugation for their full thickness. If, instead of being 

 irregularly isolated, some of these interbrachials were definitely arranged so as to form a 

 crescentic line at which the perisome commenced and then filled the remainder of the area 

 up to the tegmen, we should have substantially the structure of Forbesiocrinus. Comparison 

 of the figure of Lithocrinus divaricatus (PL XX, fig. 3a) will show that this is not a long 

 or improbable step. Thus the perisome in such forms as these may have been set with 

 thicker plates than usual, deeply imbedded in a ligamentous mass as thick as the corrugated 

 edges of the plates with which it was connected. 



This difference in the thickness of the margin is a matter of minor detail. That the 

 structure was of an unsubstantial nature, pertaining to the tegmen rather than to the dorsal 

 cup, there can be not the slightest doubt. It is only necessary to refer to the various figures 

 of Taxocrinus on Plates LIII, LV, LVII, LVIII, LIX and LX to show what the general 

 arrangement was. There we have the same crescentic margin of the interbrachial plates, 

 and in figure 3a of Plate LVII may be seen the same condition that we would have if the 

 perisome were preserved in our specimens from Tournai. By these the difference in number 

 of interbrachial plates between the new specimens and De Koninck and Le Hon's type, as well 

 as between different areas in the latter, is also explained. Being supplementary plates, form- 

 ing from time to time to fill up spaces due to the enlargement of the calyx, they vary in differ- 

 ent specimens according to age; and they may vary even within the same specimen, as 

 appears in the last mentioned figure 3a, which shows 5, 6 and 9 plates in the three exposed 

 interradii. This further appears by the condition of the margins of the axillary primibrach 

 in figures 8 and 9, PL XXIII, where in one case the sutural face extends part way up the 

 plate, and in the other all the way ; the latter would indicate interbrachials to the height of 

 IBr 3 , the same as in De Koninck and Le Hon's specimen. 



The longitudinal lines on the ventral side of the ray appear upon many of the isolated 

 plates figured, and are very conspicuous on the detached ray in Plate XXIII, figures le, d. 

 The same structure is found in Taxocrinus (PL LVII, fig. 6c), and probably throughout 



