no DISCOVERY REPORTS 



The largest specimen, St. 123, has R 95 mm., r 34 mm., br 40 mm. In shape it is 

 intermediate between Sladen's figs, i and 2, pi. 48, and considerably larger than fig. i 

 which has R 66 mm. The specimen from St. 181 has R 70 mm., r 27 mm., br 28- 

 29 mm., and is like fig. i except that the abactinal surface is much inflated and the 

 pseudopaxillae more widely spaced. The example from Marion Island (R 49 mm., 

 r 20 mm.) is closely similar to Sladen's fig. 3, pi. 48. 



It is difficult to accertain whether Acodontaster cretneus Koehler is Ludwig's species, 

 for the latter was based upon an obviously immature and inadequate specimen (R 

 18-5 mm.) collected by the ' Belgica ' in 450 m., 71° 18' S and 88° 02' W. It is figured by 

 Koehler, 1912, pi. 3, figs, i and 3, and 1920, pi. 47, fig. i. 



Koehler's cretneus, ranging in size from R 14 to 23 mm., is probably the young or a 

 depauperate form of elongatus. In his description of Acodontaster granulif ems, Koehler 

 emphasizes its similarity to cremeiis. Now, as detailed below, granidiferus seems to be 

 one of the manifestations of a race of elongatus which lives on the Falkland Plateau. The 

 similarity of cremeus and granuliferiis is explicable on the assumption that they are 

 corresponding formae of two races of elongatus. 



Pseudontaster stellatus Koehler from 64° 32' S, 97° 20' E, no fathoms, was collected 

 in the same haul as one of the specimens of Acodotitaster cremeus. The species is based 

 upon a single specimen having R 56 mm. The description of the species is therefore 

 only a description of a specimen, which I believe to be the adult of Koehler's cremeus, 

 and a form of elongatus with slightly narrower than typical marginal plates and some- 

 what shorter rays. The importance which Koehler places upon the inflation of the disk is 

 difficult to explain since such variations are of no greater significance than a full stomach, 

 swollen gonads, or an artefact of preservation. For instance, it may be produced by 

 immersing a living specimen in fresh water. The Discovery specimen from St. 181 has 

 an inflated disk. Koehler makes no mention of Acodontaster elongatus in his description 

 of stellatus, since he was convinced that Pseudontaster constituted a natural group. 

 Therefore the type of Acodontaster was naturally not in competition with stellatus. 



The type of Acodontaster capitatus (Koehler), having R 20 mm., was taken in 254 m. 

 near Adelaide Island (67° 43' S, 70° 45' 42" W), while 6 other specimens were dredged 

 in 354 and 318 fathoms, at 66° 50' S, 142° 6' E and at 66° 55' S, 145° 21' E (oflF Adelie 

 Land). The largest of these has R 55 mm., r 16 mm. It is rather obviously near A. 

 elongatus, the principal diff'erences being that capitatus has slightly slenderer and longer 

 rays, narrower superomarginal plates, and coarser granules than the typical form. It is 

 probably a high antarctic forma oi elongatus. See Koehler, 1912, p. 82, pi. 6, figs. 5, 8, 

 9, II ; 1920, p. 195, pi. 44, figs, i-io; pi. 45, figs. 5, 6. 



Type locality. Not indicated by Sladen. Probably off Cumberland Bay, Ker- 

 guelen Island, 127 fathoms, greenish volcanic mud. 



Distribution. Circumpolar, Antarctic, north to Kerguelen, Heard, Marion Islands, 

 and South Georgia. 



