308 THE INFLUENCE OF INANIMATE STJKEOUNDINGS. 



all the other conditions of life favourable for its living and re- 

 producing its kind. 



Currents themselves must no less have a dividing ac- 

 tion in some cases. It is well known that all floating ob- 

 jects, such as drift-wood, leaves, trees, &c., gradually drift to 

 the edge of the stream, even though they may have fallen into 

 the middle of it. Every navigator is familiar with the pheno- 

 mena resulting from this, and knows that the western and 

 eastern lim its of the Gulf-stream are both indicated by a broad 

 band of accumulated sea- weed, wood, leaves, and other objects. 

 This tendency of the current to clear itself — or clean itself— -is 

 stronger in proportion to its rapidity and strength. Hence, 

 objects torn by a stream flowing between two islands from the 

 one lying to the left of it, could be borne to that on the right 

 side only under specially favouring circumstances ; and vice versd, 

 those brought from the right could never, or very rarely, be 

 carried to the opposite side. Thus a mixture of the faunas of the 

 two islands might be hindered, or at any rate rendered extremely 

 difficult, simply by the action of the current flowing between 

 them. Only those free-swimming animals which might be 

 able to overcome the mechanical resistance of the current to 

 which they would be exposed in their attempt to cross it, would 

 be in a position to escape its influence. That this action of the 

 current is theoretically inevitable cannot be disputed ; still, 

 the question may of course be raised as to whether actually it 

 often comes into play. 



Certain phenomena attending the distribution or migration 

 of animals do in fact leave no room for doubt that this dividing 

 action may often be detected, above all in marine currents. We. 

 have already met with a few examples in previous sections. 

 When we were considering the striking circumstance that the 

 islands lying close to Africa have a quite different fauna from 

 that of the neighbouring continent, we mentioned this as a factor ; 

 for that fact was intelligible only on these grounds, and we 

 pointed out, on the one hand, that the stream flowing from 

 Europe, on the north, was, from the course it takes, able to in- 

 troduce a quantity of European forms into these islands, while, 

 on the other any species of animals carried off from the African 



