230 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



row of marginal cii-ri (figs. 152, p. 221, 165, p. 229, and 175-180, p. 231), to a large 

 conical or columnar plato twice as long as broad at the base, with nearly or quite 

 a dozen rows of cu-ri, wliich may be irregularly placed, arranged in crowded alter- 

 nating rows, or situated in definite columns (figs. 192, p. 237, 203, p. 239, and 20S- 

 216, p. 241). 



iVlmost always the sockets on the centrodorsal to which the cirri are articu- 

 lated are confmed to that organ; but in the calomctrid genus OreomHra, and in 

 certain of the species of the related genus Neometra, the proximal portion of the 

 sockets of the peripheral cirri commonly is shoved forward onto the radials for a 

 considerable distance so that the cirri are borne partly on the centrodorsal and 

 partly on the radials. The axial canals, however, through which the axial cords 

 passing from the interior to the exterior of the centrodorsal run, is always entirely 

 within the substance of the centrodorsal, tliough it may be only just below its ventral 

 margin. 



Ordinarily the cirrus socket is plane, or is marked with a peripheral row of 

 tubercles, and is in every way comparable to the so-called syzygy in the stem of 

 the pentacrinities just beneath the nodals, and to the articulation between the larval 

 comatuhd stem and tlie developing centrodorsal (figs. 192, 194, p. 237, 203, 207, 

 p. 239, and 208-216, p. 241). But in the species of the genera of the Atelecrinidae 

 (figs. 123, p. 192, 124, 125, p. 193, 218, 223, p. 243, 227, p. 245, and 573, 574, pi. 8), 

 and in a few other forms, this syzygy is not developed, the articulation between the 

 cirri and tlie centrodorsal being of the same type as that found between the individual 

 cirrus segments (fig. 587, pi. 13), or between the columnals in the bourgueticrinoid 

 tjT^e of stem (fig. 139, p. 205), and consisting of two ligament masses, one on either 

 side of a fidcral transverse ridge. 



The dorsal pole, or apex, of the centrodorsal is always bare of cirri (figs. 146-150, 

 p. 220, 171-173, p. 231, and 191, 193, 195, 197, p. 237), and is usually flat or more 

 or less concave, though it may be slightly convex, especially in small species. Wliile 

 most commonly smooth, it may be slightly pitted (figs. 199, 201, p. 239), or studded 

 with fine spines or papillie (figs. 191, 193, 197, 198, p. 237, 203, 205, 206, p. 239, and 

 214, p. 241), oreven with large tubercles (figs. 189, 190, p. 235). In lateral profile the 

 sides of the centrodorsal are seen to be always more or less convergent tlistally, unless 

 the centrodorsal be very thin, wliile the ventral outline, as well as the outline of 

 the bare dorsal pole, is always more or less pentagonal (though occasionally almost 

 circular), and may bo sharply stellate. ITsually the sides of the centrodorsal are 

 every^vhere uniform in cliaracter (figs. 146-150, p. 220, 171-174, p. 231, 183-188, 

 p. 235, and 219-222, 224-226, p. 243), but sometimes tlie surface is broken up into 

 five radial areas by elongate-triangular bare interradial spaces (figs. 208-213, p. 241), 

 interradial furrows (fig. 123, p. 192), orstrong interradial ridges (figs. 191-194, p. 237, 

 203, 204, p. 239, 214-216, p. 241, 227, p. 245, and 558, pi. 5), which may be supple- 

 mented by similar but less prominent structures situated in the midradial line (figs. 

 203, p. 239, and 227, p. 245), in the latter case dividing the centrodorsal into 10 defi- 

 nite areas, 2, a right and a left, in each radius. The cirri may thus be evenly dis- 

 tributed over its surface (except at the dorsal pole), or may be segregated into 5 or 

 10 radial areas (very rarely occumng in a single column in the midradial line) (fig. 



