Although the geographical distribution of fresh-water fish and fresh-water shells is often surprisingly extensive and apparently capricious, this may be explained by the means of dispersal being here so varied--not only aquatic birds, floods, and whirlwinds, but also geographical changes of water-shed having all assisted in the process. Moreover, in some cases it is possible that the habits of more widely distributed fresh-water fish may have originally been wholly or partly marine--which, of course, would explain the existing discontinuity of their existing fresh-water distribution. But, be this as it may (and it is not a question that affects the issue between special creation and gradual evolution, since it is only a question as to how a given species has been dispersed from its original home, whether or not in that home it was specially created), the point I desire to bring forward is, that where we find a barrier to the emigration of fresh-water forms which is more formidable than a thousand miles of ocean--a barrier over which neither water-fowl nor whirlwinds are likely to pass, and which is above the reach of any geological changes of water-shed,--where we find such a barrier, we always find a marked difference in the fresh-water faunas on either side of it. The kind of barrier to which I allude is a high mountain-chain. It may be only a few miles wide; yet it exercises a greater influence on the diversification of specific 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 difference 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.