ASTEROIDEA i95 



hyaline portion of the spine (about half total length). Less often (Sts. 148, WS 81) 

 the three sides are concave in the middle region of the spine, greatly accentuating the 

 three-edged condition (Fisher, 191 1, pi. 116, fig. 3^, Pt. temnochiton). The edges are 

 rounded, rather than sharp. Rarely a fusiform suboral spine has a flattened tip termi- 

 nating in 3 or 4 small, sharp, points. 



The tube-feet of the proximal half of ray are more crowded than in the strictly biserial 

 arrangement, but it is probably an exaggeration to call them quadriserial. In the largest 

 specimen the crowding continues to end of ray. The same condition exists in militaris, 

 and is still more accentuated in marsippus and jordani (Fisher, 191 1, pis. 98, loi). 



The specimen from St. 148, South Georgia, measures R 37 mm., ri4mm., 

 R = 2-57r. The adambulacral combs contain usually 5 or 6 spines. Marginal mouth 

 spines 4-6; suboral spine, rather more slender than usual, 3-edged throughout the distal 

 hyaline half, and the facets of the spine concave. Normally, of course, the spine is 

 heavily sheathed, except the triagonal point. It is questionable whether even this 

 protrudes in life. The general habit is in no wise fundamentally different from the 

 Falkland Plateau specimens. 



The range of variation in the number of adambulacral and oral spines is in line with 

 that of other species where the series of specimens is large enough to yield pertinent 

 information. The variation in the shape of the suboral spine is somewhat disconcerting 

 since the extremes have been used in differentiating species. In this series of Pt. lebruni 

 we probably have indicated two formae, but there is far too little material to reach any 

 conclusion. 



This series indicates that Pt. lebruni is much more closely related to Pt. militaris than 

 either Clark (1908) or I (191 1) supposed when we compiled our "keys" to the species of 

 Pteraster. For instance, we distinguished militaris from lebruni on the difference in the 

 number of adambulacral spines which proves to be insignificant. Also I state that the 

 suboral spine is not three-edged, whereas it sometimes is. No harm is done by this 

 since the contrasted species, trigonodon, has long paxilla stalks carrying 6 spines. 



Species of the southern hemisphere which are related to lebruni are : 

 brachiatus Koehler, 1917, p. 49. pl- 7, Ags- L 2, 6; pi. 10, fig. 6. Kerguelen, 2 specimens. This seems 



to be a race of lebruni. 

 aculeatus Koehler, 1920, p. 168, pi. 38, figs. 3, 4, 5; pl- 65. %■ 6. 65° 06' S, 96° 13' E, 325 fathoms, 



I young specimen probably close to, or the same as, Pt. affinis Smith. 

 rugatiis Sladen, 1889, p. 473- pl- 74- Ags- 3- 4; pl- 77- H^- 3- 4- Between Kerguelen and Heard 



Islands, 150 fathoms. ^ 



marplatensis Bemasconi, 1937, p. 169, figs, i, 2; pl. i, figs. 1-4; pl. 2, figs. 3, 4. Argentina, 39 S, 

 57° 10' W, 45-97 fathoms. Also four other localities between 37° 40' and 38° 25' S, at 56° 30' W, 

 48-55 fathoms. 

 argentinus Bemasconi 1937, p. 173- fig- 3 ; P'- 2, figs, i, 2. Argentina, 39° 26' S, 56° 25' W, 55 fathoms. 



Pt. marplatensis Bemasconi is very closely related to lebruni and seems to me to be at 

 best a race of that species. Bemasconi states that it differs from lebruni in the greater 

 number of oral and adambulacral spines. But this number (7 or 8) is not greater than I 

 find in large specimens of lebruni. 



