134 DISCOVERY REPORTS 



Cryaster brachyactis (Clark, 1923, p. 293, pi. 11, figs, i, 2) was dredged in Algoa Bay, 

 South Africa, in 26 and 57 fathoms. It is very different from Perknaster. It has shorter 

 and distally broader rays ; a definite series of large inferomargiual plates, some of which 

 bear a conspicuous spine (shown in Clark's figure) ; remnants of large abactinal and 

 superomarginal plates which are not tabulate or paxilliform; large actinal plates, and 

 very wide adambulacral plates. The skeleton, other than the ambulacral and adambula- 

 cral systems is undergoing degeneration. The body wall is thick and leathery. The 

 species apparently represents a new genus of the Asteropidae {vide infra). 



In order to differentiate accurately the several nominal forms of Perknaster it has been 

 necessary to compare the structure of the spinelets, especially the abactinal, as well as 

 the pedicel of the abactinal plates which carries the spinelets. Whether the small 

 differences are as important as implied in the notes under the different forms only 

 future experience will determine. It has seemed desirable to carry out an analysis of the 

 most tangible or demonstrable details and leave to the future the decision whether we 

 are dealing with a wide-ranging species, having numerous geographical races (each 

 possibly with recognizable formae) or with several species, similarly subdivided. 



The difference between ftisais, with its low pedicels and broadly clavate spinelets, 

 and sladeni, with high pedicels and slender spinelets, seems obvious enough, although 

 superficially the specimens are very similar. The characteristic crowded podia of 

 aurorae seems to constitute a fundamental character, especially in conjunction with a 

 spine structure definitely distinguishable from that oi fuscus and sladeni. 



It will not be surprising if P. charcoti is found to intergrade with P. fuscus. It must be 

 recalled that the material oi fuscus is still meagre and that the species probably reaches a 

 size equivalent to the type of charcoti. 



Koehler (1917, p. 36) has described Cribellopsis based upon C. rallieri from Kerguelen 

 (low tide or very shallow water). The figures (pi. 5, figs, 6-10) suggest at once a species 

 allied to Perknaster charcoti. The actinal interradial areas are crossed from inferomargiual 

 to adambulacral plates by lines or narrow grooves, and each file of plates, between two 

 lines, corresponds to an adambulacral plate. The creases are doubtless ciliated grooves. 

 These occur in P. charcoti. Koehler says: " Les papules que j'ai signalees sur la face 

 dorsale se continuent egalement sur la face ventrale ou elles restent tres nombreuses, 

 mais elles n'empietent pas sur les lignes de separation des rangees transversales de 

 granules." This is an extraordinary state of affairs: the papulae do not encroach upon 

 the areas of separation between the transverse rows of granules. That is, they are absent 

 from the interval between the lines of plates, where they would be expected to occur. 

 Papulae are difficult to distinguish in some specimens and it is possible that Koehler 

 mistook pustules of skin, which often resemble papulae, for these organs. The matter 

 needs confirmation by examination of the coelomic surface of the actinal interradial 

 areas for pores. Of course no Perknaster has actinal papulae. 



I desire in this connection to render grateful acknowledgments to Mr D. Dilwyn 

 John for having made comparisons of Discovery specimens with the types of Perknaster 

 fuscus and P. densus, and later for the loan of a cotype of P. fuscus for examination. 



