12Q BULLETIN 82, UNITED STATES NATIONAL MUSEUM VOLDME 1 



range of material I cannot agree with his conclusions with regard to these species. It 

 seems to me tlmt all the Atlantic species of Antedon tend to intergrade with one another 

 on tlieir geographical borderlines and much more material, particularly from the north 

 of Scotland, the western Mediterranean and from off Senegambia is needed to clarify 

 tiieir limits. 



At the same time Gislim decided that the character by which Air. A. H. Clark 

 distinguished Compsometra from Antedon, namely the very spinous distal edges of the 

 pinnvdo segments in the species of Compsometra, is not sufficient to justify the retention 

 of two genera as it is somewhat variable in Antedon and A. bifida particxdarly may have a 

 similar spinous condition well developed. 



Of the seven species which Mr. Clark had included in Compsometra, six, namely 

 loveni, incommoda, serrata, iris, longicirra and parviflora are from the Indo-West Pacific 

 while the seventh, nuttingi, is from deep water in the West Indies. However, nuttingi 

 cannot be confused with the other Atlantic species of Antedon, since, besides the depth 

 at wliich it is found, the cirrus segments are greatly elongated, up to five times as long 

 as broad, and also Pj is a genital pinnide. (In fact, the last character, together with the 

 apparent absence of the species from shallow water, makes me doubt whether it should 

 have been included in Compsometra in the first place and whether it can now be placed 

 in Antedon.) Compsometra longicirra, parviflora and to a lesser extent also iris, have 

 the cirrus segments also more or less elongated, but the remaining species have them 

 short and sunilar in proportions to those hitherto included in Antedon, the length not 

 exceeding two and a half times their width. 



I had myself referred another species to Compsometra which must now be termed 

 Antedon, namely Repometra arabica A. H. Clark, 1937. 



The genus Repometra, of which arabica was the tj^e and only species, was included 

 by Mr. Clark in the subfamily Thysanometrinae, on account of the short segments of 

 P,, but the stout cirri are quite unlike the delicate ones of the Thysanometrin species, 

 most of which also come from deeper water than does arabica. 



A final addition to the genus Antedon is a new subspecies of A. (formerly Compso- 

 metra) incommoda from the Dampier Archipelago, northwest of Australia, which I am 

 caUing auslini. The type specimens were recorded by Mr. Clark in 1911 as "Compso- 

 metra sp." but in this typescript were included by hun mider the heading of C. iris. 

 I found that their cirri are more like those of incommoda and their centrodorsals are 

 much flattened, approximating m form to the discoidal ones found in the types of 

 incommoda from Port Philhp. Dr. H. L. Clark has recorded incommoda from Western 

 Australia as far north as Geraldton and it may be that his specimens are intermediate 

 with, or even referable to, austini. 



With regard to the validity of the species of Antedon in the former sense (i.e. 

 excluding those of Compsometra) Mr. A. H. Clark liad a number of comments, of which 

 those not disproved by recent work follow here [end of notes by A.M.C.]. 



The two species called mediterranea and adriatica seem to stand well apart from the 

 others. Though when typically developed quite different from each other and easily 

 recognizable at a glance, they are very closely related and undoubtedly intergrade. 



Antedon pctasus, so far as I have seen, exhibits relatively little variability, in marked 

 contrast to A. bifida, which is exceedingly variable. 



Along the west coast of Scotland and in deep water in the Irish Sea a form occurs 

 which is scarcely, if at all, separable from true petasus. Either true petasus occurs 



