132 DISCOVERY REPORTS 



rayed form from Stanley Harbour, Falkland Islands (St. 55) which is similar to the 

 specimen from the same locality figured by Koehler (1923, pi. 7, fig. 5). This South 

 Georgia specimen has very small marginals. 



The other 6 specimens are somewhat peculiar and suggest Perknaster (e.g. P. densus). 

 The rays are more pointed than in northern forms and although narrow for the genus 

 are characteristically inflated at the base. The marginals are very small and in 2 specimens 

 from St. 170 they cannot be distinguished from the other plates. Yet these specimens 

 are not separable from two taken at St. 167, South Orkneys, in which the marginals, 

 though only slightly larger than adjacent abactinal plates and not larger than plates of 

 radial region, can be distinguished by their contrast to the actinals. 



Mr D. Dilwyn John kindly compared one of the specimens from St. 170 with the 

 type of Perknaster densus. In densus, briefly, the abactinal and lateral surfaces are more 

 densely covered with spine groups, the groups and spines being more closely packed. 

 While in densus the actinal spines and groups are close enough together to obscure the 

 arrangement of the plates, this is not true of the South Shetlands specimen where the 

 groups are slightly spaced. The adambulacral armature is essentially the same in both 

 forms. Fundamentally there are 2 furrow spinelets, slightly longer in the Discovery 

 specimens. Mr John points out that the adambulacral spines of densus are less uniform 

 than shown by Sladen's figures and description. Proximally the furrow pair form a 

 somewhat oblique series and the distal member is more prominent into the furrow. Then 

 in the middle part of the ray the spines are as in Sladen's figure (1889, pi. 98, fig. 12), 

 but distally there is only owe furrow spine. This is true of the Discovery specimens. In 

 one specimen from St. 167, the furrow series remains slightly oblique all along the ray, 

 with the slightly larger distal spine more advanced into furrow, until on the outer fifth 

 of ray it, alone, stands on the margin and the other one diagonally behind it, precisely as 

 Mr John has sketched it for densus. In densus he shows a group of four other subequal 

 spinelets behind the second spine. In the Discovery specimens there may be a group of 

 3, or all 5 may form a fairly regular transverse series on the plate. In either case the 

 arrangement is that of typical Perknaster. 



The series of 13 young in the absence of adult specimens would probably be classified 

 as Perknaster rather than Cycethra. The skeleton is well knit as may be expected in the 

 young of Perknaster. As compared to P. antarcticus the abactinal spinelets are more 

 numerous (3-8 per plate) and the skin thinner. But the spinelets are very much like 

 those of Perknaster as the figures will show. Only the first few plates have two adambulacral 

 furrow spines. Over most of the ray there is but one. Back of this spine there are three or 

 four variously disposed, sometimes making an oblique transverse series of four with a fifth 

 aborad to the third or fourth spine (or the interval between them). This is the typical 

 arrangement in Perknaster. 



Five of these young and the smallest adult were taken at St. MS 71. A spine of this 

 adult (R 25 mm.) is figured (fig. 40) for comparison with the abactinal spines of P. 

 aurorae from St. MS 71 (figs. 3, yi). In the latter specimen the abactinal paxillar pedicels 

 are higher than in Cycethra. 



