igS DISCOVERY REPORTS 



spinelet is a cluster of smaller and more slender spinelets". If this character is constant 

 in capensis, then gibber, with a small tuft of slender central spinelets, is more like 

 tesselattis. In full-grown specimens of tesselattis there are 7 or 8 stout peripheral spines 

 and upward of 18 much more delicate ones in a central group. 



The colour of the 3 is much the same— dark or pale violet : capensis sometimes has a 

 dark angular ring on the upper surface on a paler ground colour. Tesselattis is usually 

 mottled with fawn colour, but may also have darker markings. Gibber seems to have 

 been dark violet grey in life. 



The figure of the oral plates of capefisis (Mortensen, he. cit., text-fig. 8) would serve 

 almost as well for gibber. In our specimen the suboral spines are slightly longer and 

 heavier, but circular in section (as in capensis). Ludwig's specimen of gibber had 6 

 marginal spines. In capensis the number [fide Clark) varies from 5 to 7. 



Generic position of Pteraster gibber. The close similarity of gibber to tesselattis 

 and capensis, which are regarded as Pteraster, naturally excludes it from Retaster. Clark 

 (1923, p. 298) has restricted Retaster to cribrosiis von Martens and insignis Sladen. He 

 designated cribrosiis as the type species, under the impression that this formality had 

 never been complied with. When Perrier instituted the genus^ he listed Pteraster 

 capensis and Pt. cribrosiis without fixing either as type. Sladen^ was the first to give the 

 group a standing but neglected also to designate a type. In 1919^ the present writer 

 chose Perrier's first species, Pteraster capetisis, for the type of Retaster. At that time, 

 this seemed to be a perfectly safe procedure since Perrier (1875, p. 382) had seen the 

 type of Gray's capensis and stated: "Par la structure de leur tegument dorsal, soutenu 

 par un reseau a large mailles formees de ligaments unissant les epines qui le soutiennent, 

 et par leur grande taille, ces deux especes {capensis and cribrostis) s'eloignent des autres 

 Pteraster, et peut-etre faudra-t-il creer pour elles un genre special. 



"Un echantillon unique desseche au British Museum. — Du cap de Bonne-Esper- 

 ance." 



Unfortunately, since Retaster has been formally attached to Pteraster capensis Gray it 

 becomes, for the time being at least, a synonym of Pteraster in its widest sense. 



The genus heretofore known as Retaster may appropriately be called Etiretaster, type 

 Retaster insignis Sladen. Included species : the type and Pteraster cribrosiis von Martens. 

 The reasons for choosing insignis for the type are : there is a definite, well-preserved type 

 specimen from Challenger St. 189,* Arafura Sea, 25 fathoms; Sladen's illustrations are 

 excellent and detailed. Since Pt. cribrosiis has been designated (though erroneously) the 

 genotype of Retaster, it seems inadvisable to employ it again as the genotype of Etiretaster. 



Type locality of Pteraster gibber. Challenger St. 311, off entrance to Smyth 

 Channel, 52'' 45' 30" S, 73° 46' W, 245 fathoms, blue mud, bottom temperature 46° F. 



Distribution of Pteraster gibber. Strait of Magellan and appropriate depths on 

 both coasts; 27-500 m., blue mud, fine sand, pebbles, shells. 



1 Noiiv. arch. Mas. Hist. Nat. 2^ ser., 1878, I, p. 56. 



- Sladen, 1889, p. 477. ^ Fisher, 1919, p. 460. 



* Designated as type locality, Fisher, 1919, p. 461. 



