32 



the normal sea water blue, and that is an accurate enough 

 description for recognition purposes. The actual colour of 

 the cold water is a pale muddy or sandy colour with a tinge of 

 green, though, when dipped up in a vessel, it is seen to be much 

 more transparent than the blue water. The muddy and blue 

 water, described by Mr. M'Lachlan at Stumpnose Bay, on the 

 west coast, agrees with the condition at Knysna, and here 

 also dead fish (Sardines) " in tons " were found in the muddy 

 water. It is not mentioned, however, that the muddy water 

 was particularly cold, and the presence of hundreds of dead 

 sea Duikers is a puzzling element. It cannot be readily sup- 

 posed that these birds died of cold, nor can it be imagined that 

 they died of hunger in the presence of so many dead fish. The 

 only explanation that would seem to fit in with the other 

 factors, is that they died in some way from eating the dead 

 fish, for which however there is no evidence. 



Other cases in which the coldness of the water seems to have 

 played a part are those near East London, 26th September, 

 1903, and at Mossel Bay at about the same time (shortly before 

 the 2oth). In the former the temperature of the water as 

 observed at East London fell somewhat abruptly from 62° F. 

 on the 7th to 50° F. on the loth (the minimum for the year), 

 and thereafter gradually became normal. The actual tempera- 

 tures from the 7th to the 12th were 62°, 57°, 56°, 50°, 51°, 56°. 

 On the other hand, we have the evidence of the trawler that, 

 on the 8th to the nth, while trawling 18 miles east of East 

 London, the water was so warm that it attracted special 

 attention. This would, of course, be at some distance from the 

 shore, so that, while the cold water was inshore, the water 

 further to sea was warm. The presence of warm water in the 

 pools at Keiskama may have been due to the shallowness of 

 the water and its heating up after the occurrence. Tempera- 

 tures were also observed at Port St. John's further to the 

 North, and these showed a variation of 1° only during the 

 same period, so that the cold water was rather local. In the 

 occurrence at Mossel Bay the presence of a " cold mist envelop- 

 ing the Bay " was adduced as evidence of the existence of cold 

 water, " perhaps due to an iceberg." Icebergs are not known 

 to navigators to occur within such a distance of the South 

 African shore as would have this effect. 



In the Table Bay case (1837) and Cape Receife (1886) there 

 is no mention of cold water, nor of any peculiarity in the colour 

 of the water. 



The coastline affected is of particular interest. Such 

 occurrences do not seem to be known in Natal, and are con- 

 lined, so far as records go, to the coast extending from East 



