OCEANIC ISLANDS. 189 



too, is it to find that America, when discovered, pos- 

 sessed no horses, a family which geology tells us 

 originated here, and which did exist until recent 

 times, and for which the country was most admira- 

 bly fitted, as has been since proved. Though some- 

 what surprising to find this so, it is not to be won- 

 dered at that we are ignorant as to the cause of this 

 sudden extinction, for we know almost nothing of 

 the surroundings of the past. Difficulties there are, 

 then, and some of them very puzzling ones ; but, as 

 a rule followed in a vast majority of cases, the dis- 

 tribution of animals on our continent is such as can 

 be explained according to the supposition that spe- 

 cies, genera, families, etc., have all arisen at central 

 points, and have from here dispersed in various 

 directions, giving rise constantly to new forms. 



Oceanic Islands. 



All the cases thus far considered are animals on 

 continents. But there are many tracts of land in 

 the ocean which are not and never have been con- 

 nected with the mainland. Most of them are vol- 

 canic in their origin and are situated, at least this 

 is true of the so-called oceanic islands, not less 

 than five hundred miles from the neighboring 

 continents. Now, such islands have their own in- 

 habitants. The first question is, of course, as to 

 where their original inhabitants came from. Only 

 two answers to this question are possible. Either 

 they were especially created for these islands, or 

 they have been derived from the neighboring main- 

 lands. Upon the former supposition we should not 



