Accidental means of dissemination, such as those to which I have just alluded, and others to which we might appeal, will generally account, and with much presumptive truth, for the many exceptional cases which present themselves, during our investigation into the effects of natural barriers, as visible in the distribution of the Annulose races, on the earth's surface. I say "exceptional cases," because any one who has laboured practically in mountain tracts cannot have failed to recognize the marked difference which is often displayed by the insect population on opposite sides of some alpine chain; whilst he whose lot has been cast amidst island groups, will have become even more conscious than the former of the permanency of those impediments which have been placed (in this instance by the broad arms of the mighty ocean) as checks upon a too rapid system of diffusion. But if the sea and mountain ranges, when of a sufficient age _in situ_, are amongst the most effectual of Nature's barriers against the self-dispersion of the animate tribes; it follows that, if the two could be (as it were) _united_, we should have found the greatest obstacle which physical conditions can ordinarily present against the wandering capabilities of the latter. The question therefore arises,--Is it possible for them to _be_ so joined? Undoubtedly it is: and hence we arrive at the conclusion, that a _mountain island_ should afford us the _minimum of size, as regards the areas its species have overspread_, which any country is able to furnish. Madeira is a mountain island,--its highest peaks rising, although resting on so small a base, to an altitude of more than 6000 feet. Yet it is only partially a case in point; for, although it was a mountain mass, and perhaps a very elevated one, when its endemic beings made their first appearance upon its surface, we have already intimated that it has become isolated _since_ that epoch: so that, whilst _one_ of the natural barriers against dispersion which it involves (namely, mountain ridges) may be considered as primary; the _other_ (to wit, the sea, as it now obtains) has played, as an agent of obstruction, but a secondary part. Still, there is good reason to believe that the ancient tract of which it is a portion was broken up at a comparatively early date after the creation of those peculiar organic forms which found their birthplace within its bounds; and that, consequently, the latter could not have wandered far (if we except those species on which unusual powers of diffusion were bestowed) when the land of passage began to give way. Hence, even the sea, in this particular instance, partakes almost of the character (no less than the mountain heights) of an original impediment; and Madeira therefore may be safely quoted as an example in which two barriers, of a primary nature, are united; and where, consequently, we may anticipate those ultra phaenomena of _areal limitation_ upon which we have been just commenting.