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



relatively much larger than in that type (PI. I. figs. 6d, Sd), so as to reveal a large 

 number of the internal ridges which separate the inner openings of the cirrus-canals. 

 Five of these, interraclial in position, are much more prominent than the rest, as is well 

 seen in the figure (PI. I. fig. 6d). The basal grooves on the ventral surface of the centro- 

 dorsal are scarcely more distinct than they are in Antedon eschrichti (PI. I. fig. 8d). 

 But on the other hand the dorsal surface of the radial pentagon bears a very well-defined 

 basal star, of which there is rarely any trace in that species (PI. I. figs. 6c, 8c). The 

 rosette lies deeper than it does in Antedon eschrichti, and the basal rays connected with 

 it are unusually stout. This is most noticeable in their isolated condition (PI. I. 

 figs. 7, a, b; compare PL III. fig. 2, and PI. IV. figs. 4-6, all equally magnified). In some 

 cases their distal ends appear on the exterior of the calyx, as seen in PI. I. fig. 6a. This 

 figure too shows the difference between the articular faces of the radials in Antedon 

 eschrichti and Antedon antarctica respectively. Their slope is more uniform in the 

 latter species, as there is much less of an angle between the upper and lower parts of each 

 face than in Antedon eschrichti (PI. I. fig. 8a), and the consequence is that more of the 

 large muscular fossag is visible in the ventral aspect of the calyx (PI. I. figs. 6b, 8b). 

 The ridges which separate them from the lower pair of fossae are much more horizontal 

 than in Antedon eschrichti, so that the two pairs of fossse are of very unequal size 

 (PI. I. fig. 6a, 8a). 



A detailed comparison of the two outer radials and of the lower brachials in the two 

 species respectively reveals a number of similar points of difference between them ; and 

 though they are so very closely similar in habit and in general appearance, as also in the 

 conditions of their existence, there can, I think, be no question that they are distinct. 



3. Antedon australis, n. sp. (PI. XXVI. figs. 4, 5 ; PL XXVII. figs. 14-20). 



Specific formula — A. — . 



Centro- dorsal hemisjiherical, thickly covered with about fifty cirri. These have 

 twenty-five to thirty joints, nearly all of which are longer than wide. The later joints 

 are laterally compressed, and their dorsal edges project considerably beyond the bases of 

 their successors, thus giving rise to a strong spine in the last few joints. The young 

 cirri round the dorsal pole resemble the mature form, but have fewer joints ; those round 

 the margin may have thirty smooth and elongated joints which ouly develop spines 

 quite late. 



First radials just visible ; the second short and nearly oblong, but little incised for 

 the axillaries, which are broadly pentagonal or triangular with a slight backward pro- 

 jection in the middle of the base. 



Ten arms, of somewhat overlapping joints, but not tubercular at the base. The arm- 

 joints after the second syzygy are shortly triangular, gradually becoming quadrate, but 



