SYSTEMATIC REPORT 43 



open ocean far from land, and it is remarkable that the species has not been taken in the equatorial 

 waters between the northern and southern zones from which records are known. 



The vertical distribution is significant and may explain to some extent its horizontal distribution. 

 With one exception all the specimens of L. spinosus have been taken at relatively shallow levels over 

 great depths, in the North Atlantic over the great North American trough which surrounds the 

 Bermudas and in the South Atlantic over the Brazilian trough. The Discovery specimens were taken 

 in oblique hauls of 300-0 m. and 260-0 m. over depths of 5095 m. and 5272 m. respectively. Fage 

 states that the records from the northern hemisphere prove that adults of this species inhabit by pre- 

 ference levels of 330-700 m. over depths of about 5000 m., while juveniles inhabit somewhat higher 

 levels. In the southern hemisphere the Dana captures were made in open vertical hauls, so that it 

 was not possible to ascertain the actual depths at which the animals were living, but one haul was made 

 at night from only 10 m. to the surface. 



Both the hauls in which this species was captured by 'Discovery II' were taken at night and at 

 somewhat shallower depths than most of the previous records. It may be that these animals perform 

 vertical migrations to higher levels during hours of darkness as is the habit of many species of 

 mysids. 



Lophogaster rotundatus Illig, 1930 



(Fig. 3 G-K) 



191 1 Lophogaster typhus var. Tattersall, p. 120. 



1930 Lophogaster rotundatus, Illig, p. 405, figs. 



1939 Lophogaster rotundatus, Tattersall, p. 224. 



1942 Lophogaster rotundatus, Fage, p. 15, figs. 



Occurrence : 



St. 440. 21. ix. 30 (day). East of Durban, 1050-950(^0) m., 1 imm., 14 mm. 



Remarks. Tattersall (1911a, p. 120) recorded an ovigerous female of 20 mm. from south of the 

 Saya de Malah Bank in the western Indian Ocean as Lophogaster typicus var. He pointed out certain 

 differences between it and the published descriptions of L. typicus but, with only a single specimen 

 available, he did not feel justified in instituting a new species for it. Illig (1930, p. 405) founded a new 

 species, L. rotundatus, for a female of 17 mm. captured in the Zanzibar Channel and Tattersall 

 (1939, p. 204), when recording three specimens of this species captured by the 'John Murray' 

 Expedition, stated that on re-examination he found that his L. typicus var. belonged to the same species. 



A single immature specimen captured in a vertical haul off the east of Durban by the ' Discovery ' 

 very closely resembles L. rotundatus, but shows some few differences which may be due to immaturity. 

 The central rostral spine extends only slightly beyond the lateral ones and extends barely to the middle 

 of the third segment of the antennular peduncle whereas in Illig's type it is longer than the lateral 

 ones and extends beyond the distal margin of the antennular peduncle. The integument of the cara- 

 pace is devoid of tubercles and the postero-lateral angles are evenly rounded as in the type. The 

 antennal scale has the same very broad shape as in the type but, although the animal is immature, there 

 are four teeth on the outer margin, whereas Illig's much larger specimen had only three (Fig. 3H). 



The lobe on the inner side of the distal margin of the third segment of the antennular peduncle is 

 so short as to be almost obsolete and its anterior margin is very finely serrated (Fig. 3 G). 



The uropods are relatively much shorter and the endopod stouter in this specimen than in the type 

 and the armature of the telson is somewhat different. Illig figured two very small lateral spines and 

 two very strong large spines at the distal end near the apical spine. Tattersall recorded that in his 

 specimens there were three small spines along the lateral margins and only one large lateral spine at 

 the distal end near the apical spine. The Discovery specimen has exactly the arrangement described by 



6-2 



