84 THE VOYAGE OF H.M.S, CHALLENGER. 



It is curious that these protecting anambulacral plates ou the genital pinnules of 

 Antedon acoela should be so largely developed/ while those which cover the inter- 

 palmar areas of the disk are comparatively small and irregular in character (PL LV. 

 fig. 5). 



In many Comatulce, however, the disk is very closely plated, both in the ambulacral 

 and in the interambulacral areas. The plates of the latter are mostly small, and rarely 

 pierced by more than one water-pore ; while the ambulacra are generally marked by an 

 irregular double row of transversely oblong plates, as in Antedon angusticalyx (PI. LV. 

 fig. 6). But these are sometimes barely distinguishable from the anambidacral plates, 

 and the whole set encroach very much upon the peristome, so that it is scarcely visible in 

 the dry state, as shown in Antedon basicurva^ (PL LV. fig. 7). This is still more 

 marked in Antedon acoela (PL LV. fig. 5). Both the ambulacral and the anambulacral 

 plates are palisade-like in form, as the former are in Pentacrinus decorus (PL XXXIV. 

 fig. 2); and they are very much crowded, so that the course of the food-grooves can only 

 be made out with difliculty even in spirit specimens, while the peristome is frequently 

 entirely invisible. 



The disks of the three Antedon species just mentioned are very much incised between 

 the ambulacra, so that they are markedly stellate in form. The arrangement of the 

 coiled digestive tube is consequently much less complex than in large disks Like those of 

 Antedon eschrichti or of Pentacrinus, w^hich have the rays united by perisome so as to 

 increase the capacity of the cup ; while the interradial spaces are filled up with connec- 

 tive tissue which supports extensions of the digestive tube. 



In some species of Antedon with an incised disk the anambulacral plates are somewhat 

 squamous, with a tendency to overlap one another. This is the case, for example, on the 

 disks represented on PL LV. figs. 3, 4. They probably belong to Antedon nmltlradiata^' 

 having been dredged at Cape York in an isolated condition, together with entire indivi- 

 duals of this species. The edges of the interpalmar areas rise rather sharply towards the 

 ambulacra, which are marked by strong ridges with indications of a median groove visible 

 upon their upper surface. The food-groove beneath is really comparatively deep, with its 

 edges plated somewhat regularly and turned in towards one another. This is very 

 marked in the immediate neighbourhood of the peristome, which is thus comj)letely closed. 



1 The position of the plates which protect the genital glands as regards the pinnule-joints, and the very regular 

 appearance which they frequently present, have led me to think that the so-called " rudimentary pinnules " of 

 Cijathocrinus loiic/imatms, Angelin, may possibly he of the same character, and not ambulacral side plates more largely 

 developed than usual (see chap. iv. pp. 62-66). A comparison of Angelin's figures of these structures as seen from the 

 side and from above (Tab. xsvi. figs. 4b, 4c), with the corresponding tows of the protecting plates in Antedon acoda 

 (PI. LIV. figs. 2, 3), shows a remarkable similarity in their number and general aiTangement. In the recent forni, 

 however, these plates are on the jiinnules, while those which they ajjpear to resemble in the Pala?ozoic Cyathocrinus 

 are on the arms. But as these bear no pinnules, and must therefore have themselves contained the genital glands, the 

 dift'erence between the two structures is not so great as it would seem to be at first sight. 



2 The specific formula of this type is — A . 10 . — 3 ^Jie specific formula of this type is — A . R . 3 . 3 . — . 



