406 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.43. 



Carpenter's (Antedon) quadrata is a synonym of this species. 



Very small specimens of this form have P3 more like P^ than like 

 P2, and the borders, sometimes also the dorsal surface, of the ossicles 

 of the IBr series and lower brachials are prominently spinous; with 

 increasing size the spinosity gradually disappears, and P3 gradually 

 comes to resemble P2 instead of P^. 



HATHROMETRA PROLIXA (Sladen). 



1. Spitzbergen; Romer and Schaudinn; Cat. No. 4391. — One 

 specimen with arms 120 mm. long; the longest cirri are 45 mm, long 

 with 38 segments; the apical cirri are 15 mm. long with 22 segments 

 which are proportionately shorter and more "dice-box shaped" than 

 those of the peripheral cirri. 



2. Spitzbergen; Romer and Schaudinn. — One specimen with an 

 arm length of about 100 mm.; P^ is 17 mm. long with 31 segments, 

 Pg is 7 mm. long; the distal pinnules are 16 mm. long. The cirrus 

 segments appear to be a trifle shorter than usual. 



3. Spitzbergen; Romer and Schaudinn; Cat. No. 4386. — One speci- 

 men. 



4. Spitzbergen; Romer and Schaudinn; Cat. No. 4387. — Two speci- 

 mens. 



5. Spitzbergen; Romer and Schaudinn; Cat. Nos. 4389 and 4390.— 

 Two small specimens. 



6. Spitzbergen (Stations Nos. 36, 37, 39); Romer and Schaudinn. — 

 Numerous specimens. 



7. North Atlantic Ocean; Norwegian North Atlantic Expedition. — 

 Two specimens with an arm length of about 90 mm. 



The species of the genus Hathrometra are among the most difficult 

 of the recent crinoids to determine on account of their excessive fra- 

 gility. The chief differential characters are found in the cirri; but 

 these are commonly, one might almost say usually, lost. Moreover 

 the cirri of each individual are very variable in size, those about the 

 periphery of the centrodorsal being twice as long as those about the 

 dorsal pole, or even longer, and all possible gradations occurring. 

 Only the peripheral cirri are of value in specific determination so that, 

 unless these are preserved, identification becomes a matter of the very 

 greatest difficulty. The number of these diagnostic cirri, as distin- 

 guished from the nondiagnostic subperipheral or polar cirri, is never 

 very great; in large species like H. prolixa there are scarcely more 

 than from half a dozen to a dozen, though in smaller species like H. 

 sarsii the number of cirri available for systematic use is considerably 

 greater. 



Aside from the very great length and the large number of compo- 

 nent ossicles in the diagnostic cirri of H. prolixa, these are somewhat 

 abruptly larger than the other cirri and do not grade into them imper- 



