50 BULLETIN 8 2, UXITED STATES NATIONAL MUSEUM VOLUME 1 



in An/yrometra find hemispherical in Euantedon). However, he describes the centro- 

 doi-sui of .1. cru^pa, the tj^pe species of Argyrometra, as hemispherical. Also in two 

 species of Euantedon, tahitiemis and sinensis, no complete P, is known. Since different 

 species e.xhibiting; a considerable range in number of segments of P, have been mcluded 

 in single genera elsewhere in this subfamily— notably 12 to 23 in Toxometra and 8 to 49 

 in Antedon (12 to 28 m yl. serrata alone) — I do not think that a difference of about 

 5 or 6 segments can be used as a character of generic significance. In 1928 Dr. H. L. 

 Clark described a species from South Australia, paucicirra, which he placed in Euantedon 

 though it has only 10 to 12 segments in P,. If the diagnosis of Euantedon is broadened 

 to accommodate this species then the only character which may be of sufficient weight 

 to distinguish the two genera is the very large number of cin-i, LX or more, in the 

 relatively small known specimens of Argyrometra crispa and mortenseni. However, 

 in tlie typescript Mr. Clark has included in Argyrometra the species Iridometra exquisita 

 which has only XIj-L cirri at an arm length two-thirds again as great as in the type 

 specimens of the two other species. It seems to me better to refer exquisita to the 

 rediagnosed Euantedon. 



In the following key the inverse proportions of the proximal pinnules in the two 

 species of Argyrometra have necessitated their separate inclusion. Although the key 

 splits off Antedon petasus from the rest of the genus, the different proportions of Pj 

 relative to P3 seem to provide the best means of separating Euantedon and Mastigometra 

 from Antedon. 



KEY TO THE GENERA OF ANTEDONINAE 



a". P3 not of the same length and character as the succeeding pinnules. 

 6'. Pa the longest and stoutest pinnule on the arm. 

 c'. Distal ends of the cirrus segments not overlapping the bases of those succeeding; dorsal edge 

 of the outer 4-6 cirrus segments about as long as the proximal border (Philippines to north- 

 west Australia; 0-510 meters) Toxometra (p. 51) 



c'. Distal ends of the cirrus segments more or less produced and overlapping the bases of those 

 succeeding; outer cirrus segments much longer than the proximal width. 

 d'. Smaller (arms up to 60 mm. long) with not over 17 cirrus segments; cirri less numerous, 

 XX-LI (rarely over XL) (southern Japan and the Bonin Islands to Tonga and northern 

 Australia, and westward to the east coast of Africa and Mauritius; 0-728 meters). 



Dorometra (p. 61) 



cP. Larger (arms up to 80 mm. long) with 16-33 cirrus segments; cirri more numerous, up 



to LX (rarely less than XL) (Phihppines, southern Japan and the Lesser Sunda Islands; 



69-146 meters) Eumetra (p. 77) 



b'. Pj not the longest and stoutest pinnule on the arm. 



c'. Pi, P2, and Pj similar and of approximately equal length, with at least 13 segments, usually 

 about 18 (southern Japan to Timor and the Admiralty Islands; 80-731 meters). 



Iridometra (p. 86) 



c'. Pj much longer than both Pj and P, though similar to the latter (Andaman Islands to southern 



Japan; 55-274 meters) Andrometra (p. 81) 



a'. Pj of the same length and character as the following pinnules. 

 6'. Pj a little larger than Pi. 



c'. Pi \\-ilh 18 to 35 segments, not more than twice as long as broad (South Africa and southern 



Japan; 0-47 meters) Annametra (p. 91) 



c'. Pi with 8 to 12 segments up to four or five times as long as broad (New Zealand; 110-175 



meters) Argyrometra mortenseni (p. 98) 



y. Pj smaller than Pi, often only half aa long, 

 c'. Pj distinctly smaller than Pj. 

 d'. Pi with 10 to 21 segments. 



