REPORT ON THE HOLOTHUMOIDEA. 139 



the calcareous ring, to which it is closely united, and opens by a minute orifice into a 

 very narrow duct, which lies close to the inwardly-directed side of the calcareous ring, 

 and turning round the anterior portion of this latter, passes into the radial ambulacral 

 vessel (PL XLIII. fig. 7, y). That part of the narrow duct which is situated at the 

 inwardly-directed side of the calcareous ring, and which presents a small expansion, gives 

 off two pairs of branches (PL XLIII. fig. 7, c) the posterior being longer than the anterior 

 one; these branches are the true tentacular canals. 



In Oneiropkanta and Deima, &c, the tentacular cavities reach a considerable size, 

 and are supported posteriorly by the calcareous ring (PL XLIII. fig. l,k); anteriorly 

 they pass into the branches or processes of the terminal part of the tentacles, but if there 

 be no such, as in Uyodcemon raaculatus, Lcetmogone, &c. (PL XXXIX. fig. 4, and 

 PL XLIV. figs. 11 and 14), the water- vascular system gives off a number of branched 

 or unbranched csecal prolongations within the thick sole-like terminal part. The ten- 

 tacles of the Elasipoda never possess ampulla?. The five radial ambulacral vessels, 

 which run backwards along the longitudinal muscular bands, on the inner side of the 

 ambulacral nerves, and which sometimes, as in Lcetmogone ivyville-thomsoni, reach a 

 considerable width (PL XLII. fig. 1, e), commonly give off very short lateral branches 

 which enter pedicels and processes ; in some cases there are no such lateral branches 

 present, but the pedicels and processes communicate directly by a minute aperture with 

 the ambulacral system. A transverse section of an ambulacrum shows very distinctly 

 that the radial ambulacral vessels are separated from the neural canal by a firm, 

 apparently homogeneous elastic layer of connective tissue (PL XLII. fig. 9, e). It 

 is rare to meet with Elasipods, in which exist true ampullae, freely depending into 

 the peritoneal cavity, notwithstanding which two kinds of such ampullae may be 

 observed — the simple and the branched. The simple ampullae have been found only in 

 Uyodcemon rnaculcdus (PL XLII. fig. 3), where they attain a considerable size, from 

 10 mm. to 15 mm., thus exceeding the Polian vesicle in size, and where they are only 

 present in communication with the dorsal ambulacra. The branched ampullae, which are 

 met with in Orphnurgus asper (PL XLI. fig. 3, b), and Pannychia moseleyi, really belong 

 to the processes and consist of small vesicles, which send out a number of very short, but 

 comparatively wide csecal prolongations. The pedicels appear regularly to lack such 

 ampullae, at least I did not find them, excepting, as in Orphnurgus asper, in communi- 

 cation with the large ambulacral cavities or rooms (PL XLI. fig. 3, c), which will be 

 described further on. 



Often no ampullae nor anything corresponding to them can be discerned, and this 

 seems to be especially the case in the Psychropotidae, where, however, canals very 

 commonly occur remarkable for their length and width, which He within the perisoma 

 and are in direct communication with the pedicels and processes. In this family 

 the broad, flat, characteristic brim which surrounds the body is penetrated by a 



