OCTOPODIDAE 393 



Dimefisions {in mm.). 



Dorsal mantle, length ... ... 36 



Head, width, % mantle length ... 86 



Mantle, width, % mantle length ... iii 

 Arms, % total length ... ... 57 



Suckers, diameter, % mantle length 8-3 



Web, index ... ... ... 60 



The body is globular, and the head distinctly narrower than the body. It is, however, 

 rather broader than is usual in charcoti. The eyes are distinctly larger than in most 

 examples of the latter. The arms (see above) 

 are very short for an adult Octopod, being 

 amongst the shortest recorded. The suckers 

 are small and close set. The sectors of the web 

 are, I think, subequal, C and D being slightly 

 deeper than the others. The web is very large, 

 and attains the great depth of 60 per cent of 

 the arms. The surface is rather a rich, light 

 purple dorsally, becoming paler ventrally. It 

 is covered dorsally and laterally by a number 

 of close-set warts. These are very curious, and Fig. 16. Thaumeledone gunteri. Radula. (The 

 confer on the animal a rather mossy appear- remains of the degenerate second laterals are 

 ance. On the sides of the web and on the arms 



they tend to be circular and bubble-like. On the anterior surface of the web, head and 

 body, they are branched and irregular, and it is here that they are seen to stand out 

 very clean-cut from the surface. They remind me of the similarly clean-cut warts in 

 Octopus pallida, though they are not stellate in gimteri. It is possible that they may be 

 like those of a form of charcoti. Joubin (1906, p. 6) says that in the latter some of the 

 warts were probably branched. From the figure {loc. cit. pi. i, figs. 1-2) it is quite evident 

 that, in charcoti, the warts on the head and body are in contact with each other, while 

 in gunteri they are quite separate. In Joubin 's later figure (i9i4[?], p. 36) and in one 

 of the ' Terra Nova ' specimens the warts are quite separate but, in the former, they are 

 granular, and in the latter, they are mammiform, both very unlike those in gimteri. 



The funnel is short, narrow and pointed. The funnel-organ is V V-shaped. The limbs 

 are rather slender and pointed at each end. They remind one of those oi E. " aurorae" 

 (Berry, 1917, fig. 14). There are five to six filaments in each demibranch, a very low 

 number. The ink sac is absent. The radula is degenerate and represented by a simple 

 unicuspid rhachidian, the mesocone of which is low and stout. There are faint traces of 

 admedian teeth and of an oblong second lateral with a low cusp. 



Remarks. This species has a superficial resemblance to E. charcoti. For a long time 

 I considered that it should be treated as a well-marked sub-species of the latter, especi- 

 ally since only a single specimen is available, and that a female. The profound diff^erences 

 in radula, etc., were then discovered. The degeneration of the radula and loss of the 

 ink sac, length of the arms, depth of the web, number of gill filaments, the sculpture and, 



