392 



/ 



DISCOVERY REPORTS 



in some from Station 142). This ridge or keel is absent in the specimen from Station 

 MS 63. The radula (Fig. 15) has a rhpchidian with a long stout mesocone, the height 

 of which exceeds the base. There are, on some teeth, faint traces of entocones; but the 

 tooth is to be regarded as unicuspid. The admedian is small, narrow and has a low, 

 blunt entoCCne. The second lateral has a rather narrow and shallow base with a low, 

 heavy subterminal cusp. The third laterals are small, much curved and have wide bases. 

 The colour in all these specimens varies from dull olivaceous purple to brown. 



Remarks. I am far from certain as to the status of all the forms which I have placed 

 in this species. Perhaps the issue will be narrowed a httle, if we realise that, of all the 

 previously described species, its closest relationships are with E. charcoH. It clearly has 

 no connection with the smooth Antarctic forms adelieana and albida, from both of which 

 it differs in many characters. 



Similarly it is very unlike turqueti. The squatter phase has some likeness to E. har- 

 rissoui and, indeed, from the key given by Berry {loc. cit. p. 14) one might think that the 

 two forms would prove to be identical, except for the difference in the form of the 

 funnel-organ. Moreover, the "dull clouded slaty grey" colour, alluded to by Berry, is 

 very often found in polymorpha. On the other hand, the form of the funnel-organ and 

 radula, the shape of the eyes, head and body and the greater depth of sector A of the 

 web (20-23 per cent in harrissoni, 30-27 per cent in polymorpha) are all characters which 

 make it impossible to identify the two forms. 



There remains E. charcoti, and with that form the resemblance is nearer. As far as 

 the bodily proportions, arms, suckers and gills are concerned, the two species do not 

 differ in any material respect, except of course in so far as the form ohlonga is much 

 narrower. The web of the latter, but not of form affinis, is as deep as that of charcoti. The 

 sculpture in the more extreme forms (e.g. the type) of charcoti is, of course, very dis- 

 tinctive; but there are specimens in which it is more granular and like that oi poly- 

 morpha. The following characters, which seem to be regularly associated, are, to my 

 mind, good differentia: the form of the funnel-organ, hectocotylus and radula, and the 

 regular presence of a prominent and well-developed keel (either absent or seen merely 

 as a poorly developed ridge in charcoti). It must be borne in mind that, in the array 

 before us, there are a certain number of individuals which exhibit various combinations 

 of the characters of the two species, and may be hybrids. The species are to be 

 recognised on account of the higher frequency of certain character-associations. 



Genus Thaumeledone, n.gen. 



The radula is very degenerate, being reduced to little more than the rhachidian teeth. 

 There is no ink sac. The funnel-organ is double. The gills are very much reduced. Type 

 of the genus: Eledone brevis, Hoyle. 



Thaumeledone gunteri, n.sp. (PI. IV, fig. 3.) 



St. 158. 21. i. 27. 53° 48' 30" S, 35° 57' 06" W. 401-411 m., rock. Large dredge: one ? (Brit. 

 Mus. 2) [Type]. 



