PART 5 A MONOGRAPH OF THE EXISTING CRINOIDS 831 



of each arm of cubensis are wall-sided, while they are not so in balanoides. He also em- 

 phasized the greater elaboration of the prominences on the centrodorsal in cubensis. 

 H.L. Clark unfortunately made no comment on the shape of the proximal brachials of 

 his specimen, but as it did retain the bases of some cirri, unlike the holotype of cubensis, 

 he was able to contrast these with the stumps remaining on the Blake holotype of 

 balanoides, which was presumably from station 151. Carpenter's description and 

 figures (and more particularly those of Hartlaub) of this specimen show that there are 

 three or four short segments at the base of the cirri followed by much more elongated 

 segments. H. L. Clark describes the cirri of the Atlantis specimen as having only the 

 first segment short, the second being already twice as long as broad and the following 

 ones progressively longer. He comments that they are much more slender and cylindri- 

 cal than in balanoides, where they are apparently rather flattened laterally, at least at 

 the base. If the enlargements given in Hartlaub's captions to his plate 14 are accurate, 

 the height of the centrodorsal in the holotype of balanoides is about 4 mm. and only a 

 little larger than that of the Atlantis specimen. Such a small difference in size would 

 not be sufficient to account for the more slender cirri of the Atlantis specimen, and I am 

 inclined to believe that it is H. L. Clark who is correct in thinking that two distinct 

 species of Atelecrinus do occur in the West Indian area. However, with no material at 

 hand but the single Challenger specimen of balanoides, which lac.ks the cirri, I can come 

 to no conclusion about this problem. The differences in the ornamentation of the cen- 

 trodorsal, in the shape of the proximal brachials, and particularly in the cirri, may well 

 prove to provide valid distinctions, but until more material is available, I am leaving 

 cubensis (pourtalesi) in the synonymy of balanoides, where Mr. A. H. Clark placed it. 



ATELECRINUS HELGAE A. H. Clark 



Atelecrinus helgae A. H. CLARK, Fisheries, Ireland, Sci. Invest., 1913, pt. 4, p. 4 (description; Helga 

 sta. CXX; comparisons). Unstalked crinoids of the Siboga-Exped., 1918, p. 267 (in key; range) ; 

 The Danish Ingolf-Exped., vol. 4, No. 5, Crinoidea, 1923, p. 44 (range [including that of an un- 

 recorded specimen]). GISLEN, Zool. Bidrag Uppsala, vol. 9, 1924, pp. 43, 47, 53 (obliquity of 

 articulations), p. 83 (additional data); fig. 12, p. 45; figs. 25, 26, p. 50; fig. 91, p. 81; fig. 117, p. 

 93. MORTENSEN, Handbook of the echinoderms of the British Isles, 1927, p. 22 (diagnosis; 

 localities); fig. 10, p. 23. KOEHLER, Lea chinodermes des mers d'Europe, vol. 2, 1927, p. 133 

 (listed). GISLEN, Ark. Zool., vol. 19, No. 32, 1928, p. 13. 



Diagnostic features. This species very closely resembles A. balanoides, from whicli 

 it seems to differ only in being smaller and in having the centrodorsal sharply conical 

 and only very slightly longer than the basal width. 



Description. The centrodorsal is sharply conical, 2.5 nun. in diameter at the base 

 and 3 mm. long interradially, measured along its side [probably about 2.75 mm. in 

 vertical height A.M.C.]. The cirrus sockets are arranged in 2 regular very closely 

 crowded columns in each radial area, 3 or 4 to a column. On either side of each cirrus 

 socket is a subtriangular thickening, the swollen fulcral ridge, which is produced out- 

 ward and slightly downward (distally) so that the centrodorsal has a strongly serrate 

 profile. In the midradial line these processes form a zigzag line running down the 

 center of the radial area. In the iuterradial lines they lie on either side of a very narrow 

 slitlike furrow. In the basal portion of the controdorsal there are 5 prominent in- 

 terradial ridges which run from the produced center of each basal downward, becoming 

 narrower, sharper and lower, disappearing between the proximal cirrus sockets. 



The cirri are XXXV, all lacking. 



