If the Fuegian flora was treated in the analogous way (and this would incidentally show how far the Cordillera are a high-road of genera), I should then be prepared far more easily and satisfactorily to understand the relations of Fuegia with the Auckland Islands, and consequently with the mountains of Van Diemen's Land. Moreover, the marvellous facts of their intimate botanical relation (between Fuegia and the Auckland Islands, etc.) would stand out more prominently, after the Auckland Islands had been first treated of under the purely geographical relation of position. A triple division such as yours would lead me to suppose that the three places were somewhat equally distant, and not so greatly different in size: the relation of Van Diemen's Land seems so comparatively small, and that relation being in its alpine plants, makes me feel that it ought only to be treated of as a subdivision of the large group, including Auckland, Campbell, New Zealand...
I think a list of the genera, common to Fuegia on the one hand and on the other to Campbell, etc., and to the mountains of Van Diemen's Land or New Zealand (but not found in the lowland temperate, and southern tropical parts of South America and Australia, or New Zealand), would prominently bring out, at the same time, the relation between these Antarctic points one with another, and with the northern or Arctic regions.
In Article III. is it meant to be expressed, or might it not be understood by this article, that the similarity of the distant points in the Antarctic regions was as close as between distant points in the Arctic regions? I gather this is not so. You speak of the southern points of America and Australia, etc., being "materially approximated," and this closer proximity being correlative with a greater similarity of their plants: I find on the globe, that Van Diemen's Land and Fuegia are only about one-fifth nearer than the whole distance between Port Jackson and Concepcion in Chile; and again, that Campbell Island and Fuegia are only one-fifth nearer than the east point of North New Zealand and Concepcion. Now do you think in such immense distances, both over open oceans, that one-fifth less distance, say 4,000 miles instead of 5,000, can explain or throw much light on a material difference in the degree of similarity in the floras of the two regions?
I trust you will work out the New Zealand flora, as you have commenced at end of letter: is it not quite an original plan? and is it not very surprising that New Zealand, so much nearer to Australia than South America, should have an intermediate flora? I had fancied that nearly all the species there were peculiar to it. I cannot but think you make one gratuitous difficulty in ascertaining whether New Zealand ought to be classed by itself, or with Australia or South America--namely, when you seem (bottom of page 7 of your letter) to say that genera in common indicate only that the external circumstances for their life are suitable and similar. (315/2. On December 30th, 1844, Sir J.D. Hooker replied, "Nothing was further from my intention than to have written anything which would lead one to suppose that genera common to two places indicate a similarity in the external circumstances under which they are developed, though I see I have given you excellent grounds for supposing that such were my opinions.") Surely, cannot an overwhelming mass of facts be brought against such a proposition? Distant parts of Australia possess quite distinct species of marsupials, but surely this fact of their having the same marsupial genera is the strongest tie and plainest mark of an original (so-called) creative affinity over the whole of Australia; no one, now, will (or ought) to say that the different parts of Australia have something in their external conditions in common, causing them to be pre- eminently suitable to marsupials; and so on in a thousand instances. Though each species, and consequently genus, must be adapted to its country, surely adaptation is manifestly not the governing law in geographical distribution.