GEOGRAPHICAL DISTRIBUTION 355 



and Kerguelen regions (cf. Regan, 1914, Vol. i, British Antarctic ' Terra Nova' Exped. 

 Rep.). It appears, on the contrary, that there is a gradual diminution in the Antarctic 

 species as we travel outwards from the Antarctic Continent. There are, however, several 

 significant features of the distribution, which have an important bearing on the general 

 problem of the distribution of species in the South Atlantic. These are : (i) that one species 

 of Antarctic and one South American species are found around False Bay; and (ii) that 

 on the eastern shore of the South American Continent some Antarctic species appear 

 to have penetrated as far north as the mouth of the Rio de la Plata, but that beyond this 

 the character of the fauna changes entirely. To understand the full significance of this it 

 is necessary to consider the distribution of sponges in the South Atlantic in conjunction 

 with that in the Indian Ocean. 



In his monograph of the horny sponges (1889), von Lendenfeld refers to a number of 

 species of Euceratose sponges as being common to both Australia and the West Indies, 

 and found only in those two areas. At first sight it seems unlikely that a species should 

 be restricted to two such widely separated areas, and unfortunately this author leaves 

 no record as to where the material on which he worked may be found, so that it is not 

 possible to verify his results by actual observation. Since, however, so many of his 

 identifications have been shown to be inaccurate, his conclusions are apt to be regarded 

 as unreliable. Moreover, the extreme difficulty of identifying species of the Euceratosa 

 with certainty only serves to increase the doubt felt on this point. Recent observations 

 have tended, however, to suggest that von Lendenfeld may have been correct in this 

 case. For example, the single commercial sponge found on the Great Barrier Reef 

 of Australia seems to be indistinguishable from some of the varieties of the Bahamas 

 grass sponge, and is, so far as I am aware, not found elsewhere unless it be in 

 the adjacent waters of the Indo-Pacific. The question remains therefore whether the 

 superficial resemblance between the Barrier Reef sponge and the West Indian grass 

 sponge may not be the result of convergence rather than of actual affinity. That this is 

 no case of convergence, and that the two have a common origin, is suggested by the fact 

 that at least five species of sponge, sufficiently well characterized to be identified without 

 hesitation, are common to both the West Indies and Australia. Moreover, in some cases 

 they have been found also in the Indian Ocean and on the west coast of Africa. This 

 seems to have an important bearing on the problems concerning the distribution of 

 sponges and on the systematic study of the group. 



I have suggested (1930) that there might be some correlation between the distribution 

 of sponges and the main surface currents of the oceans, and that the study of these 

 currents might be of material benefit in interpreting the natural and well-marked areas of 

 distribution which undoubtedly exist. This arose from the study of the sponge fauna of 

 Norway, but the conclusions reached then concerning the North Atlantic appear to 

 hold true for the South Atlantic also. Assuming this to be the case, the chart of the 

 main oceanic currents is not only sufficient to show why the faunas of the Red Sea, 

 Indian Ocean and Indo-Pacific differ little from each other, but also why that of the 

 Barrier Reef of Australia should be composed almost exclusively of species belonging to 



