THE RELATIONS OF THE SPECIES 239 



uneven and misshapen. All the individuals with abnormal rostra were not recorded as 

 they were taken and examined day by day at sea. Since then 250 specimens of E. similis 

 from eight stations and 459 specimens of E. similis var. armata from thirteen stations 

 have been re-examined ; three of the former and twenty-two of the latter had misshapen 

 or deformed rostra. 



Distribution. Hansen (1910) described the variety crassirostris from the waters of 

 the Malay Archipelago; Tattersall (1925) has recorded it from off the Cape and Illig 

 (1930) from near the Cape and the Maldive Islands. 



THE RELATIONS OF THE SPECIES TO ONE ANOTHER 



I believe that nine of the ten species described in this report form two natural groups 

 and that, if it be supposed that the genus Eiiphaiisia arose in warm waters, the existence 

 of these groups and the order of occurrence of their members show that the genus has 

 colonized the cold southern waters along two distinct lines. Along one line the coloniza- 

 tion has not been completed ; although the southernmost member occurs in the ice-free 

 water of the Antarctic Zone it does not appear to breed successfully there. Colonization 

 has been completed along the other line ; every niche has been occupied to the farthest 

 south : there is a neritic species along the coasts of the Antarctic continent. 



The two groups are : 



(i) E. hafiseni. (2) E. liicens. 



E. spinifera. E. vallentini. 



E. lougirostris. E. frigida. 



E. triacatttha. E. superba. 



E. crystallorophias. 



The first of these has long been recognized as a natural group. It is Hansen's Group d} 

 In 19 10 and 191 1 he divided the genus into four groups as follows: 



Group a. Species with two pairs of lateral denticles on the carapace. No dorsal process on third 

 to fifth abdominal segments. 



Group b. Species with a single (rarely no) pair of lateral denticles on the carapace. No dorsal 

 process on third to fifth abdominal segments. 



Group c. Species with a single pair of lateral denticles on the carapace. A protruding acute 

 dorsal process on third abdominal segment, but without any dorsal process — at most with a minute 

 denticle {E. mucronata) — on fourth and fifth abdominal segments. 



Group d. Species with a single pair of lateral denticles on the carapace. A well-developed dorsal 

 process on third abdominal segment and conspicuous dorsal denticles or processes on fourth and 

 fifth segments. 



Hansen (191 1) wrote: "the first and the fourth (Groups a and d) are well separated 

 from the others, while the second and third {b and c), separated exclusively by the 

 existence or non-existence of a dorsal process on the third abdominal segment, are 

 somewhat badly defined, because two species show individual or local variation as to 



1 Hansen included in his Group d E. spinifera, E. longirostris and E. triacantha. E. hanseui was described 

 later by Zimmer who recognized that it belonged to the same group. 



