REPORT ON THE CRINOIDEA. 81 



Although the principal diagnostic character on which Perrier established Eudiocrinus 

 atlanticus thus turns out to be due to an erroneous method of nomenclature, he men- 

 tions a subsidiary one which merits more attention. For he finds that " L'JE. 

 atlanticus se distingue egalement de ces dernieres especes par le nombre et la grandeur 

 des organes, si repandus chez les Crinoides, nomnies corps sp>hcriques ou saccules. 

 Les saccules manquent aux E. japonicus, et E. Semperi ; ils sont petits et rares 

 chez YE. varians." 1 In this respect, therefore, the Atlantic species is sharply 

 distinguished from both Eudiocrinus semperi and Eudiocrinus japonicus, which resemble 

 it most closely in the structure of the skeleton ; while they have over twenty-five cirrus- 

 joints, of which there are only fifteen in Eudiocrinus atlanticus. 



The mutual relations of the five species of Eudiocrinus may, therefore, be expressed 

 as follows : — 



I. First two brachials united by syzygy. First pinnule on the second brachial, . 1. indivisus, Semper, sp. 

 II. First two brachials united by a bifascial articulation. 



A. First pinnule on the second brachial, . . . . .2. varians, n. sp. 



B. First pinnule on the fourth brachial. 



1. Sacculi absent. Twenty-five or more cirrus-joints. 



a. Disk plated. First brachials nearly oblong, . . .3. semperi, n. sp. 



/?. Disk naked. First brachials trapezoidal, . . .4. japonicus, n. sp. 



2. Sacculi abundant. Fifteen cirrus-joints, . . . .5. atlanticus, Perrier. 



Eudiocrinus varians, n. sp. (PI. VII. figs. 3-7). 



1882. Eudiocrinus varians, P. H. Carpenter, Journ. Linn. Soc. Lond. (Zool.), 1882, voL xvi. p. 496. 



Centro-dorsal low, nearly hemispherical, bearing about twenty cirri in two rows 

 which leave the dorsal pole free. Two forms of cirrus occur in the same individual. — (l) 

 With two or three short basal joints, the last of which is nearly square, while the following 

 joint is considerably longer, and the succeeding ones still more so, reaching 3 mm. in 

 length. Terminal joints unknown. (2) Eight at least of the lower joints are quite short, 

 few of them being longer than wide, and that but slightly so. Remainder unknown. 



Radials partially visible. First brachials nearly oblong, inclined to be trapezoidal, with 

 small lateral processes which are the edges of the muscle-plates for articulation with the 

 radials. Second brachial also nearly oblong, with traces of a backward process into the 

 preceding joint, a pinnule on the right and a small process on the left side. The 

 following joints have somewhat unequal sides, with a pinnule on the shorter and a large 

 wing-shaped process on the longer side, which ceases on the sixth, or may go on to the 

 eighth joint. Succeeding brachials quadrate and unequal-sided, with the pinnule on the 

 longer side. The twelfth and following joints are distinctly longer than wide. Syzygia 

 in the fourth and eighth or ninth brachials ; then an interval of two to five joints 

 between successive syzygia. 



1 Comptes rendus, 1883, t. xcvi. No. 11, p. 726. 

 (ZOOL, CHALL. EXP. — PART LX. — 1887.) OoO 11 



