REPORT ON THE HOLOTHURIOIDEA. 137 



apparent, and it is of importance to note carefully that their ends, even if they do not 

 pierce the perisoma, are always closely united to and fused with the body-wall in the 

 medio-dorsal line ; no Elasipoda are found in which the madreporic canal depends freely 

 into the peritoneal cavity or is attached only to the dorsal mesentery. In the following 

 species the water-vascular system communicates with the exterior by one or several 

 pores : — Lcetmogone wyvilh-thomsoni, L. violacea, L. spongiosa, Ilyodcemon niaculatus, 

 Achlyonice paradoxa, Scotoplanes globosa, S. papillosa, S. robusta, Kolga nana, K. 

 hyalina, Parelpidia cylindrica, Elpklia purpurea, E. incerta, E. ivillemoesi, Peniagone 

 wyvillii, P. vitrea, P. affinis, Benthodytes typica, B. abyssicola, Psychropotes longicauda, 

 andbP. semperiana. I am fully convinced that only in a few forms, viz., Oneirophanta 

 mutabilis, Orphnurgus asper, Irpa abyssicola, Elpidia glacialis, and Benthodytes 

 papillifera, the madreporic canal does not open externally but is intimately united to 

 the body-wall in its medio-dorsal line ; concerning the rest of Elasipoda the material has 

 been too scanty to allow of any satisfactory researches. 



When examining that group of the Elasipoda in which the ambulacra! system does 

 not open externally, one finds that in most cases the madreporic canal terminates in a 

 larger or smaller porous tubercle or plate, one side of which is closely united to the 

 body-wall, while the other is free. In Orphnurgus, &c, the madreporic tubercle is divided 

 by the mesentery into two halves, each being fiat, slightly concave, and measuring about 

 4*5 mm. in length (PI. XXXVIII. fig. 10). The size, form, and structure of this tubercle 

 are highly variable in the different species. In Orphnurgus the network which composes 

 it is of a very solid structure, the free surface being roughened by numerous calcareous 

 papillae. In Oneirophanta the tubercle is more convex but without the solid structure 

 present in Orphnurgus (PI. XXXVIII. figs. 11, 12), and in Benthodytes papittifera it has 

 an almost globular shape. Elpidia glacialis has, on the contrary, no madreporic tubercle, 

 — according to Danielssen and Koren traces of one are present in a thin very minute net- 

 work, — and it appears that the terminal part of the stone canal ends caecally within the 

 body-wall. 



In most forms where the water- vascular system is in communication with the exterior, 

 the madreporic canal usually opens by one (PL XXXVIII. figs. 2 and 5) but not 

 infrequently by several pores. Danielssen and Koren 1 were, some years ago, the first 

 to describe, in their report on the Echinoderms dredged during the Norwegian North 

 Atlantic Expeditions, a Holothurid, Kolga hyalina, which had the madreporic canal 

 running out in a pore. At about the same time I observed the very same peculiarity 

 in several of the Challenger Holothurioidea. The pore or pores always pierce the body- 

 wall at the very place where, in the other Elasipoda, the madreporic tubercle joins the 

 body- wall, that is, in the medio-dorsal line at a longer or shorter distance from the 



1 Nyt Magazin for Naturvidenskaberne, Bel. xxv. 2, Christiania, 1879. 



(ZOOL. CHALL. ESP. PART XIII. 1881.) X 18 



