200 DISCOVERY REPORTS 



protrude above the border of a deeply emarginated web. The sHghtly emarginated web 

 figured by Sladen (pi. 77, fig. 2) is not characteristic of the species, although I find an 

 exactly similar condition on the proximal part of the rays of a specimen from St. WS 

 216; but on the distal part the spines protrude much more than in Sladen 's figure. The 

 continuous web of the first pair of adambulacrals over the outer end of oral plates is well 

 shown by Sladen in pi. 74, fig. 2. The adambulacral spines, stated by Sladen to be 5, are 

 frequently 6. 



Each oral plate has 5 or 6 marginal spines entirely without connecting web. The 

 suboral spine is slender, tapered, sharp, and normally entirely encased in membrane. It 

 is very similar to the innermost marginal spine. This latter in some instances moves 

 outward upon the actinal surface of the plate so that there are then 2 suboral spines 

 (St.WS 85). These two form with the second marginal spine (now somewhat enlarged) a 

 curved interradial series and give to the mouth angle a very abnormal appearance, since 

 the suboral spines are also more robust than in typical specimens. 



A number of specimens are carrying young which can be seen in the nidamental 

 cavity through the supradorsal membrane. In one of these which measures R 12 mm., 

 the young are very small with 2 pairs of podia and a ventral yolk sac. Already a supra- 

 dorsal membrane, showing a large central osculum, has developed (St. WS 824, 19 

 January 1932). In another specimen, R 1 1 mm., the young, about 2-25 mm. in diameter 

 and irregularly circular, have 4 pairs of podia and a large terminal "tentacle". The 

 actinolateral web is very prominent, extending as a pair of webs or fans on either side 

 well beyond the end of the radius. The remains of the yolk sack is a convex mass on the 

 mouth. The adambulacral and oral spines are differentiated but too delicate to count 

 accurately, St. 652. 



Pt. hunteri is rather typical Pt. stellifer. Koehler adduces as the principal difference 

 (the spine counts being the same) the fact that paxillar spinelets of himteri do not have 

 the symmetrical posture shown by Sladen 's figure. This is rather a naive viewpoint for a 

 veteran worker! Sladen obviously describes and figures a specimen, not a species. 

 Scarcely any structure in the entire gamut of sea star morphology is more liable to 

 modification by preservatives, or subsequent crowding, or maceration, than the supra- 

 dorsal membrane of Pteraster. A specimen from St. WS 85 exhibits the stellate posture 

 of the paxillar spinelets on two rays, but elsewhere on the abactinal surface (to quote 

 Koehler) "la disposition reguliere des pignants des paxilles est completement diff'er- 

 ente". Koehler (1920, pi. 37, figs. 4-10) gives excellent photographic figures of the 

 variations of the abactinal surface. Fig. 9 shows well the ambital fringe of actinolateral 

 spines and web mentioned above. Furthermore, Koehler (1923) records his himteri from 

 east of the Falklands. 



Type locality. Challenger St. 31 1. Off entrance to Smyth Channel, 52° 45' 30" N, 

 73° 46' W, 245 fathoms, blue mud, bottom temperature 46° F. 



Distribution. Magellanic and Falkland region to South Shetland Islands, 79- 

 500 m. ; south Indian Ocean, 64° to 66° S, 96° to 140° S, 110-358 fathoms. Circum- 

 polar, and probably divisible into local forms. 



