i 4 4 DISCOVERY REPORTS 



After we left Walvis on 6 October, a series of bottom-samples were taken up past Cape Cross 

 (WS 1082-7), an d then the bathythermograph observations were resumed, and continued up to 

 Mowe Point, with two detours (WS 1092 and 1094) inshore for bottom sampling, en route. 



In this area, and also on the 'Mowe Point' line station, numerous large shoals of fish were 

 encountered. Here again, we quote from Dr Clarke's Journal: 



9th October, 1950. In the early hours of today (0100 to 0430) the officer of the watch reported that the ship steamed 

 through large shoals of small fish, like herring or pilchards. This was in position 20 35' S., 12 55' E. to 20 10' S. 

 i2°3o'E. 



Again from 2015 to 2145 (around 19 43' S., 12 36' E.) we found ourselves passing multitudinous shoals of 

 fish. It was a dark night, and I stood and watched the shoals scatter from our approaching forefoot, one after 

 another, betrayed by a burst of greenish blue luminescence, so that the shape of each fish showed up in ghostly 

 outline. Below the forefoot, and fanning out to port and starboard, the shoals were like the showery explosion of 

 roman candles under water. 



I counted a hundred shoals (each estimated at > 100 < 1000 fish) appearing in 4 minutes 23 seconds, or 0-4 shoal 

 ( > 40 < 400 fish) per second ; the ship was then steaming at about eight knots. 



The ship was stopped in an attempt to catch some of these fish with both baited hooks and jiggers, 

 so that the species might be determined, but all efforts were unsuccessful. 



Further station work was continued with two short interruptions to mark some whales, and the 

 second survey completed on 12 October. 



Two bottom-samples were taken to the north, at WS 1 106 and 1 107, to ascertain if the 'azoic ' mud 

 extended so far north. Approaching station WS 1107, between 14.15 and 15.21 hr. on 14 October, 

 patches of blood-red water were observed, some 20-30 yards across, a bucketful was collected, and 

 microscopic examination of preserved subsamples showed that the organisms causing the dis- 

 coloration were almost certainly ciliate protozoa such as have previously been seen to cause red 

 water elsewhere (see p. 255). 



COASTAL GEOGRAPHY AND BOTTOM TOPOGRAPHY 



The west coast of South Africa stretches in a more or less N.N.W. direction from the Cape to about 

 1 8° S., then continues north and N.N.E. into the Bay of Benguela. Practically the whole length of 

 this coast is characterized by a narrow coastal belt of low-lying land gradually ascending to the high 

 interior plateaux at a distance of about 80-100 miles inland. The coastal belt is fertile in the south, 

 but gradually passes into scrub, and north of 30 S. is a truly arid desert. This desert, which con- 

 tinues to about 14 S., is known in South-west Africa as the Namib. It is almost completely waterless, 

 although heavy dews occur, and the small annual rainfall (Fig. 3) generally occurs in a very short 

 space of time, draining into the Atlantic almost as quickly. The dew is sufficient to support a very 

 scanty xerophytic flora, but cultivation is only possible up country in the more fertile valleys. The 

 presence of the desert is principally due to the cold water lying along the coast, which causes condensa- 

 tion from the lower layers of the air above it (Scherhag, 1937), and acts like a mountain range, leaving 

 the land to leeward in its rain shadow. North of the desert, in Angola, the sand gradually gives way 

 to wooded country within the range of the seasonal tropical rains. 



Although in itself arid, the South-west African coast has a very large drainage area, extending 

 across the interior to the western slopes of the Drakensberg mountains on the other side of the 

 continent. Practically all of this drainage, however, is collected in the Orange river, and other rivers 

 along the coast are only periodic torrents, during the summer rains in the south-west. 



The topography of the sea-bed is shown in Fig. 4. It appears to be affected in a large degree by the 



