140 THE VOYAGE OF H.M.S. CHALLENGER. 



great many of these canals, very wide, lying side by side (PI. XL. fig. 8, b) ; and these 

 taper towards the margin of the brim and enter the often minute tuberculate 

 pedicels or processes, which are situated in the margin. Special attention should be 

 paid to the width of these canals, of which Euphronides depressa (PL XXXIX. fig. 1) 

 affords a striking example, but at the same time it becomes evident that these canals 

 are fully analogous to the lateral branches, which the radial arnbulacral vessels in Pedata 

 give off to the pedicels and processes. 



A remarkable peculiarity in a great number of Elasipoda is the presence of large 

 arnbulacral cavities, which lie enclosed within the perisoma, and being in direct com- 

 munication with the pedicels and processes should be regarded as only continuations 

 of these. Two kinds of such cavities are present, the branched and the unbranched, 

 the former being found in Oneirophanta, Deima, Orphnurgus, Ilyodcemon, and 

 Achlyonice. In Oneirophanta mutahilis the lateral pedicels as well as the processes 

 are in communication with lame branched arnbulacral cavities. The cavities which 

 belong to the processes of the dorsal ambulacra are enclosed within the odd interambu- 

 lacrum (PI. XXXVI. fig. 4), and those communicating with the processes of the lateral 

 ventral ambulacra lie within the lateral interambulacra , while those of the pedicels, 

 principally belonging to the trivium, send out numerous branches into the lateral inter- 

 ambulacra (PL XLI. fig. 2, g). Each cavity resembles a flat room of considerable width 

 which gives off in all directions branched and unbranched, longer and shorter caecal 

 prolongations (PL XLI. figs. 1, 2). Since the cavities of the lateral ventral ambulacra 

 are closely crowded, the thick perisoma of each side of the body contains a very com- 

 plicated system of cavities and canals. Here and there the radial arnbulacral vessels seem 

 to send out a larger or smaller caecal prolongation, which does not communicate with any 

 pedicels and processes (PL XLI. fig. 1, e). 



As to the arnbulacral cavities, Deima closely resembles the above-mentioned form. 

 It has already been noted that true unbranched ampullae are present in communication 

 with the dorsal processes only in Ilyodcemon maculatus, while the pedicels of this species 

 without ampullae communicate with elongated cavities which run towards the medio- 

 ventral line and terminate in some small branched and unbranched prolongations (PL 

 XLII. fig. 4). But, even in Orphnurgus, Achlyonice, Pannychia, &c, plain evidence is 

 given that the arnbulacral cavities or vesicles of the processes do not always resemble 

 those of the pedicels. In the first-mentioned genus all the processes are in com- 

 munication with small branched ampullae, while the pedicels proceed from somewhat 

 large arnbulacral cavities, which give off a small number of large, obtuse, slightly branched 

 prolongations (PL XLI. fig. 3) ; it is, however, to be noted that these cavities send out a 

 branched freely depending ampulla of the same appearance as that of the processes, though 

 slightly smaller, thus constituting a combination of cavities and ampullae. In Achlyonice 

 the arnbulacral cavities of the dorsal processes are small, oval, and unbranched (PL XLI. 



