I40 DISCOVERY REPORTS 



At St. 887 in the pancake-ice seventy-two larvae were taken in four different net iiauls. No stage 

 earlier than Furcilia 2 was recorded, which was the most abundant form ; Furcilia 3, 4 and 5 were also 

 present. Only one of the larvae was taken in deep water, the remainder being caught in horizontal 

 nets towed at the surface or oblique nets fishing about 100 m. 



At St. 888 there were twenty-six larvae comprising Furcilia 2 and 3, with the former pre- 

 dominating. 



The small number of larvae obtained on this second visit to the ice-edge led John to believe that 

 large numbers of larvae were to be expected, not in the young newly formed pancake-ice, but along 

 the edges of the older pack-ice. This influenced the observations made on the third visit to the ice- 

 edge between Melbourne and New Zealand in late June. 



(c) Melbourne-ice-edge-New Zealand, June 1932 (Sts. 897-928). 



In June, on the third journey southward, the ice-edge was encountered in 61° S, farther north than 

 on either of the two previous occasions. The ship met light pancake-ice on the night of June 22 and 

 steamed through it in search of pack-ice, but none was found, and the pancakes became too tightly 

 packed to make a full station in darkness possible. The ship therefore steamed to the edge of the ice 

 and made a full station, no krill or krill larvae being obtained. In the morning the 'Discovery II' 

 again steamed into the pancake-ice but could see no sign of pack-ice to the south so she turned east 

 and steamed through the pancake-ice, making three oblique towing stations at four-hourly intervals. 

 No krill was taken at any of these stations nor at a station made at night at the edge of the ice. 

 Throughout the night the ship steamed eastward parallel to the edge of the ice and entered it to 

 continue the search the following morning. Oblique nets were towed and small numbers of late 

 Furcilia were taken in the upper nets. 



The i-m. and 70-cm. nets were fished on the same warp just below the surface and clear of the ice; 

 they were closed before hauling in. The i-m. net was at a depth of 2 m. and caught large numbers of 

 larvae; the 70-cm. net at 5-7 m. caught none. The nets were fished again in the same way with the 

 same results. A full station was subsequently made in the ice but no young E. superba were taken in 

 the vertical net hauls. They were caught again in large numbers in the i-m. net, and on this occasion 

 in the 70-cm. net, towed just below the ice at the end of the station. 



Krill had been taken on the previous cruises some distance from the ice in the cold westward- 

 flowing Antarctic water, but after leaving the ice on this last occasion, although oblique nets and a 

 horizontal surface i-m. net were towed every four hours until the boundary of the cold water was 

 reached, neither adult nor larval krill was taken. 



The sample of krill analysed, from the i-m. horizontal net towed at 2 m. at St. 912, consisted of 

 one-eighth of the total number, 161 6. Furcilia 4, 5 and 6 and one adolescent were taken, with 

 Furcilia 6 greatly predominating over the other stages found. 



John states that the horizontal surface hauls near the ice were enormous compared with the oblique 

 hauls. 



(d) New Zealand-ice-edge-subtropical convergence south of mid-Pacific, September 

 1932 (Sts. 942-967). 



The search for young E. superba along the ice-edge on this cruise resembled that made on the 

 previous visit, but the number taken was much smaller. Two days were occupied in the search, and 

 the night between was spent steaming eastwards along the ice-edge. Twenty-one net hauls were made 

 in the ice and along its edge; eight of them were the routine oblique hauls from 250-100 m. or from 

 100 m. to the surface and thirteen were horizontal surface hauls made particularly for the purpose of 

 catching young krill. The total number taken was small, approximately 220; only one adult was 

 obtained. 



At St. 954, 0-5 m., 108 Furcilia 6 and adolescents were taken, with the latter predominating. It was 

 observed that, contrary to the usual length distribution of young E. superba, in this instance they fell 

 into two well-defined length groups, the lower with an average about 13 mm. and the higher about 



