THE SOUTHERN GROUP 247 



All are circumpolar in range, and the order of their normal occurrence from north to 



south is as follows : 



E. lucens : all but the coldest water of the sub-Antarctic Zone. 



E. vallent/ni: all but the warmest water of the sub-Antarctic Zone. 



E. frigida : the ice-free water of the Antarctic Zone. 



E. superba : under the pack-ice and along its edge, and in the colder water of the Antarctic Zone. 



E. crystallorophias : along the coasts of the Antarctic Continent. 



The normal and the possible north-to-south range of each species is shown dia- 

 grammatically in Fig. 29. 



Figures of the copulatory organs of the five species have been grouped together in Fig. 30 so that 

 they may easily be compared. Those of E. lucens, E. vallentini and E. frigida are, despite constant 

 and unmistakable differences, so similar that it will be best to consider them first (Fig. 30 a, b, c). 



The terminal processes of all three are alike in general shape. The distal end is cleft into an inner 

 and an outer branch. In E. lucens the inner branch is much the stronger and longer; in E. vallentini 

 it is the stronger, and it may be a little longer than, or of the same length as, the outer. In E. frigida 

 both branches are flattened, not rounded as in the other two species; the innermost is the bigger, and 

 it is divided distally into two unequal lobes, the front one longer and larger than the other ; the hinder 

 edge of the outer lobe is finely serrate. 



The proximal processes of the three species are very alike. There is an angle or heel on the outer 

 side near the base. The distal part is bent inwards, and from the outer side of the bend, and a little 

 to the front, there springs a strong blade-like secondary process which is curved gently inwards. 

 On the part of the proximal process beyond it there are two thin membranous expansions, a larger 

 one behind, a smaller one with a striated appearance in front. In E. lucens the front expansion is 

 much smaller than the hinder one and does not reach so far forward (i.e. distally) ; in E. vallentini 

 it is more nearly equal in size to, and reaches nearly as far forward as, the hinder; in E. frigida it 

 reaches a little farther forward. Between the ends of the membranous expansions there is in 

 E. frigida a strong curved protuberance; a careful search shows its counterpart, a much smaller 

 tooth-like projection which is not curved, in E. vallentini and E. lucens. 



The lateral processes of the three species are very similar. There is a long nearly straight shaft, 

 the end of which is bent strongly and sharply inwards; on the outside of the bend there is a strong 

 tooth. In E. vallentini and E. frigida there is a tiny projection in the axil of the tooth on the hinder 

 side. 



There can be no doubt that E. lucens, E. vallentini, and E. frigida, having such very 

 similar male copulatory organs, are closely related species. Although the organs of 

 E. superba and E. crystallorophias (Fig. 30 d, e) are at first sight very different from theirs, 

 and from one another, I find that they share with them characters not possessed by any 

 other species of the genus. The relationship is shown by the proximal process only in 

 E. superba, by the proximal process most clearly in E. crystallorophias. 



The base of the proximal process of E. superba has no heel on the outer side as in E. lucens, etc., 

 but it is curved inwards. The distal part of the process is bent inwards and it carries two mem- 

 branous lobes, a distal one lying for the most part on the hinder side of the process and one on the 

 front side placed farther back. The latter is striated in the same way as the front lobe in E. lucens, 

 E. vallentini and E. frigida. There is no strong blade-like secondary process on the bend, but in the 

 majority of males, though not in all (p. 218), there is a spine of variable size bent forwards over the 

 middle of the foremost membranous lobe. I regard it as homologous with the larger secondary 

 process in E. lucens, E. vallentini and E. frigida. It has not been described before. 



