314 THE IKFLUENCE OF INANIMATE SUEEOUNDINQS. 



New Guinea did not rise from the sea until quite recent times, 

 still the colonising of the islands from the neighbouring con- 

 tinents might have taken place in such a way as to involve a 

 distribution such as is actually presented to us there. All the 

 larger Mammalia, being incapable of overcoming the strong cur- 

 rents prevalent there, would have been excluded from immigra- 

 tion into the newly formed islands ; only the smaller species, 

 that cling to trees, could have been carried across seas by 

 those currents ; and it agrees with this that we find all the 

 Marsupials out of Australia, as in New Guinea, the Moluccas, 

 and Celebes, belonging- exclusively to the climbing genera. And 

 that these should not have succeeded in crossing Wallace's 

 limit-line is the inevitable and very intelligible result of the 

 tendency of currents to ' clean themselves,' as before described. 

 This tendency results from the circumstance that such a current 

 is always a little higher in the middle than at the sides. Hence 

 objects floated ofi" by the right margin of a current flowing 

 through the straits of Timor or Celebes, or between Bali and 

 Lombok, could reach the left shore only under some specially 

 favourable circumstances ; they would usually remain on the 

 same side, particularly when they were passively borne along, 

 as would be the case with uprooted trees and so forth. In 

 looking at a map on which the currents in question are laid 

 down, it is at once seen that the currents flowing from Australia 

 and the southern part of New Guinea are suddenly diverted 

 from their slightly westerly or quite northerly direction to a 

 north-easterly or quite easterly flow, exactly by the very island 

 — namely Celebes — where the mixture of Indian and Austra- 

 lian forms is most conspicuous. Land animals — such as land- 

 snails — of which the transportation can only be effected by cur- 

 rents must have been subject to the same influence, and it is 

 therefore quite intelligible when we find that two islands lying 

 so close together as Bali and Lombok exhibit less similarity 

 than, for instance, Celebes and Java ; for the current that parts 

 those two little islands is so strong that it must be quite impos- 

 sible for molluscs, or other creatures that avail themselves of 

 drifting trees for their voyages, ever to pass from one island to 

 the other. On the contrary they might, under certain circum- 



