MONOGRAPH OF THE EXISTING CRINOIDS. 225 



series. Among the Thalassometridje, however, there are several species in which 

 the entire ventral brachial perisome is plated far out on the arms, more or less 

 irregular plates being present between the ventral processes of adjacent brachials, 

 and between the brachials and the side and covering plates. 



As a rule, if the disk is well plated side and covering plates are also strongly 

 developed, but there are several interesting exceptions to tliis, for instance in 

 the families Zygometridse, Himerometridse, Mariametridae, Colobometridse, and 

 Pentametrocrinidae. 



Completely plated disks are always entire, and always approach, as nearly as 

 possible, a spherical form (figs. 734-737, p. 349). Partially plated dislcs. like 

 naked disks, may assume any form. 



No one has ever studied the disk in Atelecnnus (figs. 1163, 1164, pi. 26) ; it 

 contains no deposits visible to the naked eye. 



In the species of Pentametrocrinidae no calcareous deposits are visible except 

 in Pentametroaimis semperi (fig. 1158, pi. 25) and in P. atlanticus, in which the 

 disk is well covered with a pavement of small rounded plates. In the yoimg of 

 Thaumatocrinus renovatus (part 1, fig. 117, p. 183) the perisome between the inter- 

 radials and the orals is completely covered with plates, as in the young of 

 Coinatilia iridometriforjnis (part 1, pi. 2, fig. 529) and of V omactinia mendionalis, 

 but as in these comasterids these plates appear to be resorbed. 



In the species of the family xVntedonidse the disk is naketl except in three genera. 

 In the curious genus Erythrometra (figs. 666, 667, p. 329) the disk is usually 

 naked on the ventral surface, but the lateral interradial areas bear two or three 

 columns of rounded interradial perisomic plates which are not in contact either 

 with each other or with the adjacent plates of the primary series, and there are 

 usually additional plates between the first two brachials of the arms of each pair. 

 Antedon hifida possesses a naked disk, but very frequently there are small clusters 

 of perisomic plates in the lateral interradial (interprimibrachial) spaces just above 

 the sutures between the radials, which collectively appear as single interradial 

 plates (figs. 1145, 1146, pi. 23), while in Antedon fetojius (fig. 747, p. 349), and in 

 the genus Mastigometra (fig. 750, p. 349), in addition to these comparatively large 

 scattered calcareous nodules are often found on the ventral surface of the disk in 

 the interradial areas, more particularly in the anal area. Antedon niediterranea 

 and A. adriatica (fig. 757, p. 349) have the entire perisome usually without visible 

 deposits. 



In the families Thalassometridte (figs. 738-745, p. 349) and Charitometridse 

 (figs. 746, p. 349, and 1159-1161. pi. 26) the disk is usually well protected by small 

 approximately imifomi more or less rounded plates, the sides of the ambulacral 

 grooves being always furnished with a double row of smaller irregular plates set 

 very close together. Wliile frequently in these families the plates in the inter- 

 ambulacral areas form a solid pavement which, indeed, often extends outward all 

 along the perisome of the arms, they are commonly somewhat isolated from each 

 other and, in some cases, may be widely scattered. The plating found in the 

 Charitometridse is as a rule rather more highly developed than that found in the 



