TKANSPOEIATION OF FOOD-PLANTS. 313 



the same result : namely, the production of the horse. How- 

 ever, I leave this an open question ; thus much only I think 

 may be insisted on : that in such speculations this possibility 

 should never be lost sight oi, and, at the same time, that all the 

 different causes which may have had a share in influencing the 

 distribution of animals must be fairly investigated and weighed 

 before it is possible to set up any one special method of explana- 

 tion as the only correct one to the exclusion of all others. 



At any rate it is perfectly certain that winds and marine 

 currents sometimes promote and sometimes hinder the diffusion 

 of species, both directly and indirectly. One species may be able 

 to cross a pretty sti'ong current at a sharp angle, while another 

 may be prevented by a feeble wind or current from reaching an 

 island lying very near ; certain kinds of seeds can only be trans- 

 ported by the wind, others again only by the sea ; whore, on an 

 island or a group, some pai-ticular plant is absent because its 

 sseds could not reach it, there of course the animals also will be 

 absent which depend on it for food and are monopbagous. The 

 relations thus occasioned between the faunas of islands in two 

 contiguous regions, such as the Indian and Australian, are of 

 course exti-aordinarily various, complicated and difficult to in- 

 vestigate; but this does not justify us in neglecting them. It 

 may be more convenient to argixe from upheavals and sub- 

 sidences which may be imagined in any required number, or 

 from hypothetical intervening continents whose former existence 

 can be neither proved nor disproved ; but the esisiest method is 

 certainly not always the most accurate — I may almost assert 

 that it hardly ever is. But in this particular case it does not 

 appear to me to be so exceptionally difficult to detect the rela- 

 tive conditions if we do not wilfully shut our eyes to them. 



If we now reflect more particiilarly on what has been said 

 here and in a former chapter as to the mode of action of cur- 

 -rents, the view is irresistibly forced upon us that it is possible 

 that the differences here pointed out in the distribution of 

 animals may, without exception, be explained by their agency 

 without assuming any former material connection between the 

 islands and the nearest continents. Granting that the larger 

 number of the Malayan islands and of those in the vicinity of 



