DESCRIPTION OF SPECIES jo 3 



however, the follicles project slightly from the sides of the gonad, which then approaches the condi- 

 tion typical of Styela. In fact van Name (1945) placed the species in Styela, but pointed out that 'the 

 gonads approach the Cnemidocarpa type'. Arnback (1950), however, suggested that the species 

 should be placed in her genus Ypsilocarpa, which, she says, ' agrees with the genus Cnemidocarpa as 

 regards the general structure of the reproductive organs, the ovary and testis not being separated '. 



The genus Ypsilocarpa (Arnback, 1921) was characterized as follows: 



'Branchial sac. Four folds on each side; three are well developed, and one — the second — is in a 

 rudimentary state. 



Text-fig. 36. Cnemidocarpa nordenskjoldi (Michaelsen) and C. drygalskii (Hartmeyer) : A, B, comparison of the intestinal loop, 

 ventral view; C, gonad of C. nordenskjoldi (St. WS 805); D, transverse section of the same. 



Gonads. One elongate cylindrical gonad on each side, bent into the shape of the letter U, and made 

 up of both ovary and testis. 



Atrial tentacles. The atrial tentacles are thickly scattered over the narrow velum.' 

 Arnback evidently regarded the gonads and the atrial tentacles as the best characters for separating 

 Ypsilocarpa from Cnemidocarpa, but as mentioned above found the structure of the gonad basically 

 similar. There remains only the number of the gonads, their considerable length, and their shape, 

 together with the distribution of the atrial tentacles, to distinguish the two genera, but I do not 

 consider that these characters, even in combination, are of generic value. The proper genus for 

 nordenskjoldi is therefore thought to be Cnemidocarpa. 



