214 Dr. W. B. Carpenter on the Structure [Jan. 20, 



"Wyville Thomson), from within every one of which there rises a pair of 

 non-extensile tubular tentacles. Each pair of tentacles springs from a 

 base common to it with the leaflet ; and this base communicates by a 

 single oval orifice with the tentacular canal, which lies immediately beneath 

 the epithelial floor of the furrow. (This is quite distinct from the " ten- 

 tacular canal " of previous authors, which lies below it and has no con- 

 nexion whatever with the tentacles, and which I term the subtentacular 

 canal.) The convergence of the five radial furrows towards the mouth 

 gives it when viewed from above somewhat the form of a five-rayed star, 

 the adjacent rays being separated by rounded projecting lobes of perisome 

 (Plate 8. fig. 2, m, m), beneath which (as shown at a, figs. 2, 3) is situated 

 the true mouth surrounded by a thickened annular lip. Within each of 

 the five lobes of perisome there arise from the annular lip, by separate 

 bases, a series of from eight to ten thick- walled tentacles, each of which 

 has its own orifice of communication with the interior of the oral ring. 



The true tentacular canals are the representatives of the " radial vessels " 

 described by Prof. "Wyville Thomson, in his account of the Pentacrinoid 

 larva, as extensions of the oral-ring canal, which, as will hereafter appear 

 (p. 227), is a derivative of the perivisceral cavity, marked off from it by 

 threads of connective tissue, as shown at re, figs. 10, 11. In the adult 

 Antedon, however, the thickened annular lip which surrounds the mouth 

 does not contain any distinct ring-canal, its original cavity having been 

 almost entirely filled up by threads and bands of connective tissue, and 

 by a set of caecal tubuli, as to the precise relations of which I am at 

 present in doubt. But the true tentacular canals still pass towards it, 

 and seem to lose themselves in what may be called its " interspace sys- 

 tem," this being continuous with that which has come almost entirely 

 to occupy (as shown in fig. 3, Jc) the portion of the perivisceral cavity 

 that originally lay open within the intestinal coil. 



The subtentacular canals, on the other hand, radiate continuously from 

 the neighbourhood of the mouth to the bases of the arms, each of the 

 five radial trunks divaricating into two ; but they do not give off any 

 branches or extensions whatever until they reach the arms, where alter- 

 nating lateral extensions pass into the pinnules. On approaching the 

 mouth, where the oral perisome becomes continuous with the lining of 

 the oesophagus, these canals dip downwards into the " visceral mass," as was 

 long since described by Heusinger*, though he did not trace their further 

 course in its interior. I shall presently show that they there all become 

 continuous (as at I, fig. 2) with the single " axial canal" (g), which passes 

 downwards towards the centre of the dorsal surface of the visceral mass, 

 and, through that canal, with the deeper or dorsal portion of the peri- 

 visceral cavity. And the history of the development of the subtentacular 

 canals (Plate 9. fig. 12, stc) plainly shows that they too are derivatives from 

 the upper or oral portion of the same cavity, which the progressive increase 

 * Zeitschrift fur organ. Physik, Bd. iii. (1829), p. 366. 



