NON-CONSERVATIVE PROPERTIES 199 



the anaerobic sea-bed would become magnified, and the production of hydrogen sulphide by the 

 sulphate-reducing bacteria in the bottom sediments (p. 204) would assist in the depletion of oxygen. 



The conditions found at Walvis Bay on survey I exemplify this. The previously upwelled waters, 

 warmed and mixed with the influx of oceanic water, may be considered to have produced almost 

 stagnant conditions in this region, and some few days before our arrival in Walvis Bay a small mortality 

 of fish was observed by the local inhabitants. 



The meteorological records (Fig. 35 and Table 3) show that, prior to our arrival in Walvis Bay, 

 there had been a predominance of calm and northerly winds for at least four or five weeks. It is 

 highly probable that adverse weather conditions such as these must be responsible for the occurrence 

 of such conditions of stagnation as are eventually associated with fish mortality. 



These abnormal conditions always occur in the summer (Copenhagen, 1953) during which time 

 northerly winds predominate at Walvis Bay. This fact is not shown to advantage in Fig. 5, probably 



«: imj 



Fig. 48. Predominant wind vectors at Walvis Bay during the fish mortality of the summer of 1942/3. The wind vectors are 

 plotted in the same manner as in Fig. 35. From the records of the climatological station at Walvis Bay. 



because these winds represent rather exceptional circumstances, and they do not necessarily occur in 

 every year. Other opportunities do occur, however, of comparing these abnormal conditions with 

 meteorological records. 



In January of 1943 a considerable mortality of fish was reported by Dr McConnel at Walvis Bay 

 (Brongersma-Sanders, 1948). It extended from Walvis Bay to Concepcion Bay, a distance of some 

 60 miles along the coast. 



At this time hourly meteorological observations were taken at Walvis Bay. The winds recorded 

 there are plotted (Fig. 48) in the same manner as in Fig. 35, any winds between south-west through 

 south to east-south-east being counted as positive and all other winds as negative. The negative winds, 

 mainly north to north-west, are those which would have been conducive to the abnormal conditions, 

 and Fig. 48 shows that they persisted throughout January 1943. 



A mortality of much greater extent occurred in the summer of 1924-5 (Reuning, 1925) and the wind 

 records from Walvis Bay (Fig. 49) again show that after the end of December conditions would have 

 been favourable for the production of abnormal conditions. 



One further example is a smaller mortality observed at Walvis Bay in December 1925, and through- 

 out this month the winds were either northerly or calms. 



There appears, therefore, to be a well-defined correlation, at least between the occurrence of fish 

 mortalities and the spells of calm weather or northerly winds, which create abnormal conditions in the 

 vicinity of Walvis Bay. 



