CLARK: BRITTLE-STARS. 273 



/ Basal under arm-plates noticeably swollen; no longitudinal 

 stripe on upper surface of arms. 

 g Oral surface of interbrachial areas covered by numerous 



small, thick scales wurdemanii 



g' Oral surface of interbrachial areas with scales largely con- 

 cealed under a coat of granules filograneus 



/' Basal under arm-plates not at all swollen; a conspicuous longi- 

 tudinal stripe on upper side of arm pulcher 



e' Radial shields narrow, length 2-3 times width; 12-15 series of 

 scales in interradial areas; conspicuous longitudinal stripe on 



lower side of arm septus 



d' Arms stout, only 6-9 times disk-diameter chilensis 



c' Marginal papillae of disk spinulose and very acute liitkeni 



b' Basal under arm-plates longer than wide marginatus 



a' Arms short, only 4-5 times disk-diameter; only 5-7 series of scales in 

 interradial areas where narrowest brachyactis 



Ophiophragmus wurdemanii. 



Amphiura wurdemanii Lyman, 1860. Proc. Boston soc. nat. hist., 7, p. 196. 

 Ophiophragmus wurdemanii Lyman, 1865. Illus. cat. M. C. Z., no. 1, p. 12, 



132. 

 Ophiophragmus wundermani Koehler, 1915. Bull. 84 U. S. N. M., p. 42, 



pi. 8, fig. 1, 2. 



It is unfortunate, of course, that Koehler has consistently mis- 

 spelled the specific name of this species in his brief note on its occur- 

 rence at Trinidad. His figures also are unsatisfactory as they difTer 

 from all M. C. Z. specimens of 0. wurdevianii in the indistinctness of the 

 marginal papillae and in the shape of both upper and under arm-plates. 

 The upper arm-plates in typical 0. wurdemanii are excessively^ wide 

 in comparison to their length, and are in contact with each other for 

 practically the full width. According to Koehler 's figure, the upper 

 arm-plates of his Trinidad material are only 2.5 times as wide as long 

 and are in contact for only about .(jO of their width. I did not find 

 0. wurdemanii at Tobago, nor have I collected it at Jamaica. It is 

 known from Beaufort, N. C, and from the western coast of Florida. 

 The holotype (M. C. Z., 1,119) is evidently an old individual and the 

 tentacle-pores are very large indeed. As a result the tentacle-scale 

 on the side arm-plate is often far removed from the one on the under 

 arm-plate; it is also often reduced in size and not infrequently is 

 wanting; rarely there are three tentacle-scales, the extra one being 

 on either the side arm-plate or on the under arm-plate. 



