234 SMITHSONIAN INSTITUTION 



and distal views of first iBr, PL XXIV, figs. 50, b), No. 5 sets in among these plates at the 

 same level as on the brachials (see distal view of iBr 2 , PL XXIV, fig. 6 and PL XXI, fig. 1). 



By giving attention to the articulations or sutural markings at the sides, one can identify 

 the position in the ray of any isolated plate from the radial to the axillary primibrach, and 

 approximately that of those of the second and third orders. A number of these are shown 

 in their proper succession on Plate XXIV, but it must be remembered that they are not from 

 the same individual, having been collected at random in the debris ; therefore they are rela- 

 tively disproportionate in size. In the same way from isolated plates found at an equivalent 

 horizon at Button Mould Knob, Kentucky, I have reconstructed part of a ray shown in 

 figures 1 ia, b on Plate XXIII, proving the existence here of a species essentially similar to 

 F. nobilis. 



In Forbesiocrinus agassizi from the Upper Burlington limestone the brachio- 

 interbrachial union is a modification of No. 2 ; there is a crenulated margin inclosing a large 

 fossa, which is not continuous but is broken up by a number of small facets within it 

 (PL XXV, figs. 3c, 5/, 6e) ; and this continues to the higher brachials, where the inter- 

 brachials by gradual diminution in size merge into the perisome (PL XXI, fig. 2). The same 

 (unpaired) articulation is found on the distal face of the radial (PL XXV, fig. 5?), which 

 does not agree with the theory above suggested of the homology of the last two primibrachs. 

 The compound radial having disappeared in F. agassizi (which has only 2 IBr), the articula- 

 tion on the distal face of R ought to be regularly double. We also find an imperfect articu- 

 lation in this species at the distal face of the axillary IBr 3 and proximal face of IIBr^ which 

 was an immovable plate suturally united to its fellow in the pair. 



The surfaces of articulation and union in the later species of this genus, with their more 

 solid interbrachial structures, are also somewhat different in detail from those of Forbesio- 

 crinus nobilis. In the brachio-interbrachial union the fossae are interrupted by several small 

 facets tor apposition of the solid surfaces ; that is, there was more attachment of calcareous 

 parts and less of ligamentous masses and therefore the calyx was less pliable. The muscular 

 articulations are also not so well defined (see PL XXV, figs. $e, 6d). In Forbesiocrinus 

 saffordi, a very large species marking the culmination of the genus at the close of the Warsaw 

 epoch, the loose articulation, No. 2, is not found, but the mode of union from the basals up 

 was by means of interrupted grooves and ridges, and small modified fossae (PL XXXI, 

 fig. 6c). There was still further diminution of mobility in this species by reduction in size of 

 the ligament fossae and increase of the facets for attachment of calcareous surfaces 

 (PL XXXI, figs, yb, c, e, \ob). Yet this is a matter of minor detail, the difference in plan 

 not being great. 



In Forbesiocrinus nobilis we can see the remnant of a dorsal nerve canal in the arms, 

 which is found in but few of the genera. It is shown in the articular face of many brachials 

 in the form of a median groove running from the ventral side to the point within the plate 

 where the dorsal canal should be, often becoming a distinct cavity (PL XXIV, figs. 10a, b). 



The foregoing description and discussion has been based entirely upon specimens of 

 the Belgian species from Tournai. It is now interesting to note that a form remarkably 

 similar to it has been identified from isolated plates in various localities in America at the 

 equivalent horizon, viz., the base of the Lower Carboniferous. At several places in the 

 Knobstone of southern Indiana and Kentucky these plates have been found bearing the same 

 tuberculous ornamentation, and showing the same peculiar articulations. From perfectly 

 weathered specimens I have been able to reconstruct part of a ray from the radial to the 

 axillary IBr, with the plates in their proper relative position as determined by the character 

 of the lateral faces (PL XXIII, figs. 11a, b). Here can be seen the sutures with deep fossae 

 for the interlocking of the interbrachials up to the middle of the IBr, where the corrugated 

 margin for attachment of less solid structures sets in. Distal and proximal faces of a radial 

 and primibrach are also shown (figs. 12, 13). Similar plates have been found in equivalent 



