246 DISCOVERY REPORTS 



One of the strongest differences between them is that E. spmifera has hepatic spines 

 while E. longirostris has not. It is surely significant that E. hanseni should have hepatic 

 spines like E. spinifera and that E. triacantha should, like E. longirostris, have none, but 

 that both should have a mound-like projection on the carapace similar to that from 

 which the spines arise in the other two. It is significant too that it is an end member of 

 the series that differs most from all the others — that it is E. triacantha that has no post- 

 ocular spines and a copulatory organ most unlike those of the other three. In the cha- 

 racter of the structures on the second segment of the antennular peduncle there is an 

 appearance of a progression or trend which corresponds with the order of occurrence 

 of the species. 



THE SOUTHERN GROUP 



My second natural group, which I shall call the Southern Group, comprises E. liicens, 

 E. vallentini, E. frigida, E. siiperba and E. crystallorophias and has not before been 



l/l 



u 



13 



Ul 



SUB-TROPICAL CONVERGENCE 



ANTARCTIC CONVERGENCE 



- EDGE OF PACK-ICE 



ANTARCTIC COAST- LINE 



o 



OL , 

 <liJ 

 HZ 

 ZO 



<N 



m 



10 



LJ 



z 

 o 



N 



o 



a: 

 < 



I- 

 z 

 < 



Fig. 29. Diagram showing the distribution of the species of the southern group of Eiiphaiisia in the surface 

 zones of water in the southern oceans ; the blacked in portion of each column shows the normal range of that 

 species, the entire column the possible range. 



recognized as such; E. vallentini comes from Hansen's Group c, the other four from 

 Group b. They share no obvious external characters, such as the abdominal spines of 

 Group d, to make them look like related species. I find the relationship to be shown 

 by the structure of the male copulatory organs, especially by that of the proximal 

 process. 



