_Sandwich Islands._--These are an extensive group of islands, larger than any we have hitherto considered--the largest of the group being about the size of Devonshire. The entire archipelago is volcanic, with mountains rising to a height of nearly 14,000 feet. The group is situated in the middle of the North Pacific, at a distance of considerably over 2,000 miles from any other land, and surrounded by enormous ocean depths. The only terrestrial vertebrata are two lizards, one of which constitutes a peculiar genus. There are 24 aquatic birds, five of which are peculiar; four birds of prey, two of which are peculiar; and 16 land-birds, all of which are peculiar. Moreover, these 16 land-birds constitute no less than 10 peculiar genera, and even one peculiar family of five genera. This is an amount of peculiarity far exceeding that of any other islands, and, of course, corresponds with the great isolation of this archipelago. The only other animals which have here been carefully studied are the land-shells, and these tell the same story as the birds. For there are no less than 400 species which are all, without any exception, peculiar; while about three-quarters of them go to constitute peculiar genera. Again, of the plants, 620 species are believed to be endemic; and of these 377 are peculiar, yielding no less than 39 peculiar genera. * * * * * Prejudice apart, I think we must all now agree that it is needless to continue further this line of proof. I have chosen the smallest and most isolated islands for the purposes of our present argument, first because these furnish the most crucial kind of test, and next because they best admit of being dealt with in a short space. But, if necessary, a vast amount of additional material could be furnished, not only from other small oceanic islands, but still more from the largest islands of the world, such as Australia and New Zealand. However, after the detailed inventories which have now been given in the case of some of the smaller islands most remote from mainlands, we may well be prepared to accept it as a general law, that _wherever_ there is evidence of land-areas having been for a long time separated from other land-areas, there we meet with a more or less extraordinary profusion of unique species, often running up into unique genera. And, in point of fact, so far as naturalists have hitherto been able to ascertain, _there is no exception to this general law in any region of the globe_. Moreover, there is everywhere a constant correlation between the _degree_ of this peculiarity on the part of the fauna and flora, and the _time_ during which they have been isolated. Thus, for instance, among the islands which I have called into evidence, those that are at once the most isolated and give independent proofs of the highest antiquity, are the Galapagos Islands, the Sandwich Islands, and St. Helena. Now, if we apply the method of