i66 DISCOVERY REPORTS 



The terminal segment of G. antarcticus in Eigiits's figure is certainly slightly broader at the base 

 than it is long, but measurements taken from Pfeffer's figure show a ratio of length to breadth of 

 I -16: 1 ; in other words, Pfeffer found that the terminal segment was longer than broad. The length to 

 breadth ratio of the terminal segment of Richardson's C. acutus is i-66:i. From these facts it is 

 obvious that the differences between the two species stressed by Richardson are not as marked as she 

 supposed. 



The differences between the species have also been discussed at some length by Tattersall (1921) 

 and further by Calabrese (1931), and both authors agree that the two forms represent varieties of a 

 single species and not two distinct species. 



Tattersall pointed out the inaccuracy of Miss Richardson's statements of the dimensions of the two 

 species. There is one point in connexion with his remarks which needs correction ; in referring to the 

 measurements of Pfeffer's figure, he said (p. 234) that 'the metasome is i-i6 times as long as broad'. 

 'Metasome' should read 'terminal segment'; the ratio of length to breadth of the metasome is 1-3:1. 

 Tattersall also drew attention to the differences in the proportions of the joints of the posterior limbs 

 of the two varieties, the limbs of G. antarcticus being shorter and broader than those of G. acutus. 



An examination of a number of specimens in the Discovery collections showed that the ratio of body 

 length to breadth, and metasome length to breadth did vary within a comparatively narrow range, and 

 this was most marked in the case of a specimen collected in deeper water (567 m.); the ratio of the 

 most typical ' antarcticus ' specimen, however, was linked by a number of intermediate types with the 

 'acutus' form (Plate IX, fig. 2). 



Table I, which includes measurements of specimens described by other authors as well as of some 

 from the Discovery collections, brings out this feature very clearly. In Table I the ratio of carpus 

 length to propodus length is not given, the carpus being either equal to, or slightly longer than, the 

 propodus. 



Apart from the localities cited above, G. antarcticus has been recorded by Monod (1931) from South 

 Georgia, and Nordenstam from Cumberland Bay, South Georgia (the specimen was dead and found 

 far up the shore). 



G. acutus is recorded from South Georgia (Pesta, 1928) from Gauss Station (Vanhoffen, 1917), 

 from Seymour Island, off Graham Land, and off Paulet Island, Graham Land (Nordenstam, 1933) 

 in 100-150 m. The specimens (18 in number) obtained from Seymour Island were living and col- 

 lected on the ebb-shore. 



Table i shows: 



(i) That a number of specimens described as G. antarcticus have the same body ratio as the type 

 G. acutus. 



(2) That the difference in the body ratio between a typical specimen of G. antarcticus and the type 

 specimen of G. acutus is considerably less than between the latter and the most 'acute' form. 



(3) That the metasome ratio obtained from Eights's figure is probably inaccurate, because measure- 

 ments taken from other specimens show the length of the terminal segment to be greater than the 

 breadth, indicating that the difference between Richardson's G. acutus and a typical specimen of 

 G. antarcticus is less than she supposed. 



(4) That the limb joints as exemplified by the carpus, propodus and dactylus of the last pair of 

 thoracic limbs are longer and narrower in the more 'acute' forms. 



(5) That, excluding Richardson's type specimen, the more 'acute' forms tend to occur in slightly 

 deeper water. 



(6) That both 'species' have a similar distribution, both occurring within the Antarctic Con- 

 vergence; the majority of specimens have been collected between 25 and 65° W. 



