356 Annals of the South African Museum. 



extrumarginal triangle, but the other interradials may have a tuber- 

 cle more or less well developed or may entirely lack it. The same 

 is true of the upper arm-plates. 



In view of this diversity in the arrangement, form and appearance 

 of the dorsal plates, I am very sceptical as to there being any true 

 specific distinction between cavellae and permira. It would be per- 

 fectly possible to consider the PIETER FAURE specimens representa- 

 tives of an undescribed species, marked by the central cluster of 

 spinelets and the circle of ten equal plates around the centrodorsal, 

 but in view of the locality where they were taken and the diversity 

 shown by the VALDIVIA specimens, I think they must be considered 

 cavellae. I have compared them carefully with Sladen's description 

 and figures and should have called them permira without hesitation 

 had cavellae never been described. Koehler lays stress on the absence 

 of oral shields in permira but, after examination of these specimens 

 before me, I think this is only a matter of interpretation of the 

 plate present in each adoral angle of each oral interbrachial area. 

 One of these is fairly well marked and we are all agreed in calling 

 it the madreporite, while the other four, as shown by Koehler's own 

 figures, are more or less ill-defined. In the type of permira, they 

 were so ill-defined that Sladen (1879, Ann. Mag. Nat. Hist. (5), vol. 

 4, p. 405) did not recognize their homology ; if the lowest one in his 

 fig. 5 were treated by the artist as the madreporite is, the homology 

 would be obvious! I therefore believe cavellae and permira will prove 

 to be synonyms, but until more material is available, I prefer to let 

 the South African Astropkiura continue to bear the honoured name, 

 cavellae. 



As regards the position of Astrophiura in the system, I agree with 

 Matsumoto in considering it only a highly specialized member of the 

 Ophiolepididae, and not in any sense a primitive or annectent form. 

 Sladen was carried away by the novelty of that original specimen ! 



P.F. 1909. Cape St. Blaize, N. by E. V* E., 67 miles, 90-100 fms. 

 Rough bottom. 2 specimens; adult. 



OPHIOMISIDIUM PULCHELLUM. 



Ophiomusium pulchellum Wyville Thomson; 1877, The Atlantic, vol. 2, 



p. 67. Lyman, 1882. CHALLENGER Oph., pi. Ill, figs. 1-3. 

 Ophiomisidium pulchellunt Koehler, 1914. Bull. 84 U. S. N. M., p. 32. 



It is quite natural to find this interesting little brittle-star in the 

 collection, but it is particularly noteworthy that it was taken with 



