2i8 DISCOVERY REPORTS 



peduncle. A wide lobe projects forwards from the upper surface of the second segment 

 over the third. It may be as wide as the third segment and nearly half as long, its outer 

 distal angle may be strongly produced forwards, outwards and downwards ; a keel which 

 runs the length of the inner side of the upper surface of the segment turns outwards 

 distally to end where the lobe bulges outwards above the median line of the third seg- 

 ment. The lobe may in older specimens be much smaller and shorter, its distal margin 

 straight, or it may be entirely wanting. The third segment has a high keel which may be 

 wanting in old heavily chitinized specimens. 



In the female the lobe from the end of the first segment of the antennular 

 peduncle is usually bigger than in the male; it may be wider or narrower than, and 

 anything from over one-half to less than one-third as long as, the second segment. As in 

 the male its shape is variable, but it is usually much longer on the inner side than on the 

 outer, so that its distal margin is oblique, and it is usually concave or emarginate. There 

 may or may not be a spine or projection at the outer distal corner of the first segment as 

 in the male. The lobe from the end of the second segment is usually bigger than in the 

 male, its outer distal corner more strongly produced; but it is variable. It varies from 

 being (most often) wider than the third segment to being narrower, and from being 

 more than one-half to one-quarter of its length. The outer distal corner is usually 

 strongly produced forwards, outwards and downwards so that it sometimes reaches 

 nearly as far forward, on the outer side, as the end of the third segment. There are keels 

 on the second and third segments as in the male. The lobe of the second segment and 

 the keel of the third do not appear to disappear in old heavily chitinized females as they 

 do in males. 



The abdominal segments have no dorsal spines. 



The male copulatory organ is as distinct from that of E. crystallorophias and those of 

 E. frigida, E. vallentini and E. lucens as the former is from the three latter, but it is 

 nevertheless of the same pattern as the four of them (Fig. 30 d). The proximal part of 

 the terminal process is bent nearly at a right angle to the distal part, which is about three 

 times as long. The basal part, or foot, is bigger than in E. frigida, E. vallevtini and E. 

 lucens, but the heel is not so sharp. The end of the process tapers and curls forwards; it 

 is not bifid. 



The distal two-fifths of the proximal process is bent inwards and carries two mem- 

 branous expansions, one distally, the greater part of it lying on the hinder side of the 

 axis of the process, and one on the foremost side immediately near the bend. The distal 

 end of the latter touches the former. The foremost expansion is striated, that which is 

 distal and for the most part hindermost is not. There is at first sight no secondary 

 process such as occurs in the three species already described ; none had previously been 

 seen. But it is present in the majority of males, though not in all, as a spine bent for- 

 wards over the foremost expansion. It may be as long as that shown in the figure or 

 very much smaller, one-third as long or less, or it may be entirely wanting : of thirty 

 heavily chitinized males examined it was present in the left petasmas of twenty, absent 

 from those of ten. 



The lateral process is curved at the end; it carries no tooth or secondary process. 



