SAGENOCRINIDAE 233 



3. Between radials and primibrachs. This is a peculiar kind of union which may be said 

 to be a modification of the last, producing an imperfect muscular, or perhaps only ligament- 

 ous, articulation ; there is a large dorsal ligament, and a single large ventral fossa which may 

 have been occupied by muscle, though more probably by a more or less ligamentous mass. 

 It differs from the succeeding muscular articulations in being unpaired (Pis. XXIII, figs. 12, 

 13; XXIV, figs, yb, 8a, 286). This peculiar single character has a suggestive bearing upon 

 the possible nature and origin of the radial plate. If it were in the right posterior ray of 

 Temnocrinus or any other genus having a radianal in primitive position at the base of the 

 ray, we should at once recognize in this lowest plate the lower half of a compound radial — 

 the infer-radial — and would not call it, but the next one above it, the true radial. We may 

 have the same thing here only extended into all five rays, — for every radial I have found has 

 the same unpaired distal face. Therefore the evolution of Forbesiocrinus from Temnocrinus, 

 by which the two primibrachs of the latter became what we have been calling three in the 

 former, may have been accomplished by the insertion of an infer-radial in the four regular 

 rays, thus making the radial in all five rays compound. In other words, the process of 

 equalizing the rays in this case would be not by the elimination of the radianal of the 

 Silurian type, but by the addition of its equivalent to the other rays ; that is, the increase of 

 primary plates in the ray would be by interpolation at the base, and not by addition at the 

 distal end. 



We have a similar but more complete case in the Taxocrinidae, where the Ordovician 

 and Silurian Protaxocrinus, with a radianal at the foot of the right posterior ray and two 

 primibrachs all around, is modified in the Devonian and Carboniferous first by elimination 

 of the radianal, producing Eutaxocrinus with only two primibrachs, and then by the addition 

 of a primary plate in the regular rays, resulting in Taxocrinus with three primibrachs. It 

 may be that every increase in the primary plates of the ray was by interpolation above the 

 basals, in which case only the two upper ones — the axillary and the one next below it— would 

 be strictly brachials, homologous to the two primibrachs of the primitive crinoid. Never- 

 theless, as it is only by a fortunate chance that we can observe the articulations of these 

 parts in two or three genera, and in most of them it must always remain unknown, the evi- 

 dence is not deemed sufficient to warrant a modification of descriptive terms to meet such a 

 theory. 



4. Between successive brachials throughout the ray, from .IB^ up. This is a paired 

 ligamentous articulation, with transverse ridge, dorsal ligament fossa, and two broad liga- 

 ment fossae lodging bundles of elastic ligament on either side of a vertical ridge — a modified 

 articulation comparable to the articulation on the radials in the Recent crinoids ; except that 

 here the fossae for attachment of muscles are too much reduced to be apparent in the fossil. 



5. Between distal margin of interbrachials and connecting lateral margin of brachials 

 and the interbrachial perisome already mentioned ; consisting of vertical corrugations on 

 the edges of these plates for almost their full thickness at first, but gradually shortening 

 upward and disappearing at about the end of the third order of brachials. There are no pits 

 or fossae, and the lateral faces of the plates are not angular for meeting interlocking plates, 

 as they are below where union No. 2 prevails. The change in mode of union usually occurs 

 on the primibrach next to the axillary, IBr 2 , and it may be studied in detail in the figures on 

 Plate XXIV. There it will be seen that union No. 2 extends from the base to the angle on 

 IBr, (figs, jc, 8c, 10c) ; No. 5 begins on the upper half of the last mentioned plate (fig. 10c) 

 and continues upward, being shown on axillary IBr 3 (figs, lib, 12b) ; IIB^ (fig. 14) ; IIBr, 

 (fig. 16) ; and IIIBn, or IIIBr 3 (figs. 21, 22). It may also be seen along the margin of the 

 right anterior ray in Plate XXIII, figure id. From these and other specimens there has been 

 constructed the lateral view of the ray margins shown on Plate XXI, figure 1, where the full 

 extent and relations of the two kinds of union may be seen. It will be further noted that 

 while the lower interbrachials are joined among themselves by union No. 2 (see proximal 



