REPORT ON THE CRINOIDEA. 349 



crinus kerguelensis, was obtained in various shallow-water dredgings round the coast of 

 Kerguelen Island, and was also found at 75 fathoms near Heard Island. In its general 

 facies it has a singular resemblance to Antedon eschrichti and its allies; while Promacho- 

 crinus abyssorum, from 1600 and 1800 fathoms (Stations 147, 158) is more like the 

 circumpolar and abyssal members of the Tenetta-group among the species of Antedon. 

 Two of these were associated with it at Station 147, while at Station 158 there was also 

 obtained the remarkable genus Thaumatocrinus. 



Although there are ten radials in the calyx of Promachocrinus, the symmetry of the 

 basals is only pentamerous. Five of the radials are essentially like those of Antedon, 

 with a smooth dorsal surface and two openings on the inner face, between which is the 

 shallow groove lodging the radial axial furrow. This seems to have been converted into 

 a canal by the radial process of a rosette, just as in Antedon and Actinometra (PI. I. 

 fig. 8c ; PI. III. figs. 4c, 5b) ; but I was unfortunately unable to obtain this rosette entire, 

 for the central portions of it broke away from the peripheral part which remained firmly 

 attached to the radials (PI. I. fig. lc). 



In many of the five-rayed Comatulae the interradial angles of the rosette become 

 connected with the five elements of the basal star, which are developed in the synostosis 

 between the centro-dorsal and the radials as I have explained elsewhere ; ' and these 

 basal rays lie beneath the sutures between the five primary radials (PI. I. fig. 6c; PI. II. 

 figs. 1-5, c; PL III. figs, lc, 3a, 36, 4c; PI. IV. fig. 3c; PL V. figs, lc, 5c/). In 

 Promachocrinus, however, with its ten radials (or at any rate in Promachocrinus 

 kerguelensis), there is a basal ray beneath the middle of every alternate radial (PL I. 

 figs. 1, a, c). Its inner end is broad and flattened, and extended laterally into two 

 processes which meet those of the adjacent basal rays beneath the dorsal surface of the 

 intervening primary radials (PL I. fig. lc). When these ten radials are separated from one 

 another the basal rays come away with the " interradial radials " to which they are attached 

 (PL I. figs. 2, a, b), and their impressions are left upon the inner ends of the dorsal surface 

 of the true primary radials with which they were in contact (PL I. figs. 1, 3, c). 



The isolated centro-dorsal of Promachocrinus is indistinguishable from that of 

 Antedon. Its ventral surface is marked by five grooves lodging the basal rays (PL I. 

 figs, lc, 5). But there are. only five large radial areas without any indication whatever 

 that each of these lodges portions of two additional radials, as well as its true or primary 

 one. In the large centro-dorsal of Promachocrinus kerguelensis the five interradial 

 pillars within the central cavity are very distinct, as is also the case in Antedun 

 antarctica (PL I. figs. 1, 6, d). 



In one of the three species of Promachocrinus the rays divide so as to produce twenty 

 arms ; but they remain simple in the other two species, just as in Eudiocrinus and 

 Thaumatocrinus. In both alike the first pinnule is on the second joint above the 



1 Trans. Linn. Soc. Lond. (Zool.), 1879, ser. 2, vol. ii. pp. 95-100. 



