220 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



either isolated or in grouj^s, as seen in Antedon petasus (fig. 747, p. 349) and in 

 Comatida solans. In many cases they are so frequent as to form a more or less 

 solid plating all over the surface of the disk, as is seen in the Zygometridse (figs. 

 1165-1157, pi. 25), Thalassometridte, and Charitometridag (figs. 1159-1161, pi. 26), 

 and especially in the Calometridie (figs. 735-737, p. 349). When this occurs the 

 disk is said to be " plated." 



In the young comatulids and in the later pentacrinoids there are, so far as we 

 know, no perisomic plates other than those bordering the ambulacral grooves which 

 appear almost simultaneously with the latter. 



In the j'oung of C oiriaetinia meridionalis, and probably also of other com- 

 asterids, large thin fenestrated plates forming an uninterrupted layer appear in 

 the integument during the later pentacrinoid stages, when there are as yet only five 

 cirri, the orals have scarcely commenced to imdergo resorption, the mouth is central, 

 and the digestive tube makes but a single coil. As the orals are resorbed these 

 plates also gradually disappear from the ventral surface of the disk, persisting, 

 however, until considerably later in the lateral perisome between the arms. Some 

 time after the complete elimination of these plates the perisomic plates of the usual 

 type appear. 



The plating on the disk of the young of T haumatocrinus renovatus (part 1, 

 fig. 117, p. 183), as described and figured by P. H. Carpenter, is probably of the 

 same nature. 



These plates are probably to be interpreted as homologous with the plates 

 covering the disk in certain palaeozoic types, those in the laterodorsal interbrachial 

 regions (part 1, fig. 412, p. 317), which are larger, more solid, and more per- 

 fectly formed, being the rej^resentatives of the interradials and interbrachials. 

 They exist only so long as the narrow perisomic areas between the outer borders 

 of the orals and the radial circlet are held taut and inflexible by the surround- 

 ing plates. When with the resorption of the orals and the retreat of the basals 

 these areas increase in size and movements of the external surface due to the 

 movements of the internal organs take place they soon disappear. 



They are thus in reality primary plates of the disk — ^that is to say, plates 

 which have reached the maximum development possible under the mechanical 

 limitations imposed by slight general perisomic movement. Their successors, the 

 so-called perisomic plates of the disk, as a result of the great increase in the 

 mobility of the surface of the disk during the growth of the animal, usually 

 remain at a much earlier developmental stage. 



The persistence of the plates in the interbrachial areas is due to the fact 

 that amplitude of motion here increases very much more slowly than on the disk 

 and is always very hmited. 



Comaster belli, G. novoBguinece, 0. tnultijida, and Comanthina schlegelii often 

 have the interradial portions of the lateral perisome, as well as the spaces between 

 the division series and arm bases, completely covered with a solid pavement of 

 plates by which the various primary elements are closely cemented together. 

 l^hese plates are sometimes so developed that each of them appears strongly 



