MONOGRAPH OF THE EXISTINO CRICOIDS. 339 



5-rayed types, resorbod soon after their appearance instead of developing after the 

 manner normal for the genus. 



In this connection it is most interesting to examine the figure pubUshed by 

 Mr. Frank Springer to show the probable i)rimitivo structure of ihi* anal inter- 

 radius and adjacent parts of the calyx in the whole Flexibilia t}-pe, both fossil and 

 recent. If we should carry backward to its probable inception the course indi- 

 cated by the migration of the radianal ]ilat(^ in the young of the recent comatulids, 

 we should arrive at a calyx structure identical \\-ith that shown by Mr. Springer 

 and deduced from the study of the fossil forms. From the study of the recent types 

 alone it might be argued that the figure should be slightly modified hy the r<'duj)li- 

 cation of anal x in the shape of uiterradials in aU the other interradial areas; but 

 from the data acquu-ed from the study of 6-rayed specimens, and the very evident 

 modification of all the recent tj^es in the direction of a perfect, derived from an 

 imperfect, radial symmetry, it would seem that we would be justified in considerhig 

 these four additional interadials as a later development. 



Sir WyviUe Thomson believed that the minute interradials sometimes jjresent 

 in the young of Antedon hijlda occasionally persisted and became the clusters of 

 small plates often observed in the angles of the calyx in the adult; but it is far more 

 likely, as P. H. Carpenter has suggested, that these latter arc secondary perisomic 

 plates, and that the true interradials whenever they appear are either resorbed 

 like the orals or develop into interradial radials. 



Perisomic interradials. 



In many of the recent comatuhds more or less well-defined plates are found 

 between the division series and between the first two or three brachials of the free 

 arms. These may be comparatively small and distinct, or they may be large, 

 forming a solid calcareous plating over the perisome. They are most strongly 

 developed in certam of the large very many armed comasterids, as Comaster multi- 

 fida, C. belli, C. typica and Comanthina schlegelii, and, though here restricted to 

 small areas between the bases of the IBrj, are very prominent features of certain of 

 the species oi Antedon, especially of A. moroccana and .1. diibenii (fig. 104, p. 167). 



These plates have nothing to do with true interradials of the type seen in the 

 young of Promachocrinus, TJiaumatocrinus, Comactinia, ComatiUa or Antedon, but 

 arise from a calcareous deposition witliin the more superficial layers of the peri- 

 some. These perisomic interradials will bo considered in connection with the 

 other perisomic plates and the perisomic spicules, and in the section deaUng with 

 the Pentacrinoid Larvae. 



Primary plates of the dish. 



In the young of Thaumatocrinus renovaius (figs. 115-118, p. 18.3) the surface of 

 the disk between the margin and the outer border of the orals is completely invested 

 by a pavement of small plates whicli lat(>r disapi)ear, just as does the radianal. 

 The same development of a complete but transient jilating of the disk occurs in the 

 young of Comactinia, the plates here beuig resorbed fu-st on the ventral surface of 

 the disk, and later in the lateral mterradial areas. 



