356 DISCOVERY REPORTS 



those areas. This well-defined area of distribution, the Indian Ocean and Indo-Pacific 

 area, ends abruptly in the west in the region of the Agulhas current. On the west and 

 north it is bounded by the continents of Africa and Asia respectively, and in the south 

 by the cold waters of the West Wind Drift, which is to all intents an impassable barrier 

 to migration. The eastern boundary must be ignored for the present, as our knowledge 

 of the sponges of the Pacific is so imperfect, and to avoid unnecessary complications the 

 sponge fauna of Australia, exclusive of the Barrier Reef, must be left out of the reckoning. 

 The area thus delimited, which for the sake of brevity may be called the Indian Ocean 

 area, in addition to constituting a natural and well-marked faunal area, forms a closed 

 system of circulatory currents, and it is easy to imagine that these two conditions are 

 correlative. In the same way, that portion of the Atlantic south of a line approximating 

 to the Equator constitutes a second well-defined area and forms also a partially closed 

 system of circulatory currents. To the east and west it is bounded by the continents of 

 Africa and America and to the south, again, by the waters of the West Wind Drift. To 

 the north is a similar faunal area, with its circulatory system, with which the South 

 Atlantic system is put into communication mainly by the Gulf Stream ; and such species 

 as are common to both North and South Atlantic follow a line of distribution which 

 accords with the conclusions outlined in the paper already referred to. For the moment, 

 then, it will be convenient to ignore the North Atlantic system. 



We have thus two closed systems of currents, each with its well-defined fauna, and 

 if we are to accept the hypothesis that the distribution of sponges is effected by surface 

 currents, we should expect to find a complete absence of species common to the two 

 areas. This is, as a general rule, true. A few species, not more than 2 per cent, have, 

 however, migrated from the Indian Ocean to the South Atlantic; but these excep- 

 tions are apparent only and serve to test the rule, since having once gained entrance into 

 the South Atlantic system, their distribution again follows the line of the main surface 

 currents. On the other hand, and this at first sight seems to be a much greater obstacle 

 to the development of the argument, while there is this migration from the Indian Ocean 

 to the South Atlantic, there is no evidence of a similar migration from the South Atlantic 

 to the South Pacific. This is the more remarkable since the gateway from the Indian 

 Ocean to the South Atlantic and that from the South Atlantic to the South Pacific are 

 so similar in character, both being barred by the meeting of a hot and a cold current, 

 with a considerable difference in temperature between the two. When we come to look 

 more deeply into the matter, however, we find that it is on the theory that the major 

 factor in migration is transportation by currents, rather than the limiting effects of 

 temperature, that this apparent anomaly can be explained. 



The barrier between the Indian Ocean and the South Atlantic is situated along a line 

 approximately due south of the Cape of Good Hope, and lies along the junction of the 

 Benguela current passing northwards and the Agulhas current flowing southwards. The 

 general faunas east and west of this line are markedly different, and this is almost certainly 

 due to an abrupt drop in temperature in passing from the one current to the other ; but 

 the fact that some species do pass this line shows that the barrier is not complete. The 



