86 ME. P. H. CAEPENTEE OX THE GENUS ACTINOMETRA. 



means of such a simple nature as it is in Ant. celttca (PL IV. figs. 3, 7), where it con- 

 sists solely of the five adjacent dorsal surfaces of the component radials. These are 

 somewhat elevated in the centre hut fall away towards the sides, where they are sepa- 

 rated from one another by slight furrows, corresponding in position with the five inter- 

 radial ridges on the ventral surface of the subjacent centrodorsal piece (PL IV. fig. 2, i.e). 

 In Actinometra, however, these dorsal interradial furrows are very marked ; but they are 

 not usually visible on the dorsal aspect of the radial pentagon, as they are occupied by 

 five long processes which radiate outwards from the angles of the central vacuity in 

 which the rosette lies (PI. V. fig. 3 ; PL VI. figs. 4, 9, 13, 21, S). 



The presence of these rays of the basal star introduces an element of considerable 

 complexity into the dorsal aspect of the pentagonal base of Actinometra ; and its nature 

 will be best understood if we commence with the study of its component pieces in 

 the large Act. robusta. The dorsal face of each first radial of this species (PL V. 

 fig. 12) is slightly convex, so as to fit into the somewhat depressed radial area correspond- 

 ing to it on the ventral surface of the centrodorsal piece (PL V. fig. 14, r.ar). The 

 centre of its inner margin is, as in Ant. rosacea, marked by a deep notch, which indicates 

 the position of the axial radial furrow occupying the median line of the internal face 

 (fig. 10, a.r.f.). The latter is converted into a canal by the union of its inflected edges 

 with those of one of the radial spout-like processes of the rosette (figs. 12, 13, p), in the 

 manner already described by Dr. Carpenter for Ant. rosacea. 



The central notch on the inner margin of the dorsal face thus becomes a round opening 

 (fig. 12, Q), similar to that seen in Ant. rosacea (PL IV. fig. 1G, Q). A bristle passed 

 through this opening towards the ventral side, therefore, will follow the course of one of 

 the axial radial canals, in which its lower end is concealed by the spout-like radial 

 process of the rosette (PL V. figs. 12, 13, I). On the ventral side of the rosette the 

 radial axial canal is incomplete, as the furrow on the internal face is only partially 

 bridged over by the calcareous processes which extend themselves from its sides to meet 

 the rosette ; the bristle which lies in the furrow is therefore visible here and there through 

 the openings in the network {cm) formed by the inosculation of these processes (figs. 

 10, 13, 1). This is best seen in fig. 13, which is a view of two radials from within, together 

 with that portion of the rosette which corresponds to and is united with them; and also 

 in fig. 10, which represents the internal face of a single radial, from which the portions 

 of the rosette that are normally united with it have been removed, so that the whole 

 of the internal face is exposed. The bristle I is seen to lie in the deep furrow between the 

 two raised edges of the apertures {x 1 , y) of the central canal, and to pass upwards under 

 the network extending from the ventral half of the internal face, where it follows the 

 course of the axial radial furrow and emerges on the ventral aspect of the radial. The 

 furrow in which it lies is here continued into the numerous irregular furrows of the 

 ventral face which converge towards the centre of its inner margin (PL V. fig. 11). 



Just above the dorsal surface of the radial, the axial furrow occupying the median 

 line of its internal face gives off a large horizontal diverticulum into the substance of its 

 calcareous tissue, which extends outwards for some distance between the central canal 

 and the dorsal surface of the radial (fig. 10, r.c') ; and, like the axial furrow or canal as 



