Geographical Distribution. 221 



types, where fresh-water faunas are concerned, than 

 almost any other. But why should this be the case 

 on any intelligible theory of special creation ? Why^ 

 in the depositing of species of newly created fresh- 

 water fish, should the presence of an impassable 

 mountain -chain have determined so uniformly a dif- 

 ference of specific affinity on either side of it ? The 

 question, so far as I can see, does not admit of an 

 answer from any reasonable opponent. 



Turning now from aquatic organisms to terrestrial, 

 the body of facts from which to draw is so large, 

 that I think the space at my disposal may be best 

 utilized by confining attention to a single division 

 of them — that, namely, which is furnished by the 

 zoological study of oceanic islands. 



In the comparatively limited — but in itself extensive 

 — class of facts thus presented, we have a particularly 

 fair and cogent test as between the alternative theories 

 of evolution and creation. For where we meet with a 

 volcanic island, hundreds of miles from any other land, 

 and rising abruptly from an ocean of enormous depth, 

 we may be quite sure that such an island can never 

 have formed part of a now submerged continent. In 

 other words, we may be quite sure that it always has 

 been what it now is — an oceanic peak, separated 

 from all other land by hundreds of miles of sea, 

 and therefore an area supplied by nature for the 

 purpose, as it were, of testing the rival theories of 

 creation and evolution. For, let us ask, upon these 

 tiny insular specks of land what kind of life should 

 we expect to find? To this question the theories 

 of soecial creation and of gradual evolution would 



