My dear Sir.

It is highly probable that I should have marked the Catalogue of British Plants¹ differently, if previously aware of your object. I think my leading idea was to select examples of close species or quasi-species, rather than to make the list numerically exact.

Even if aware that numerical comparison entered into your object, the Catalogue used would hardly have afforded the means for reaching much precision, because various of the genera with few species in Britain, are nevertheless large genera in Europe, or elsewhere.

In the question of close species, I should prefer the testimony of Fries before that of Hooker & Bentham united. And apart from personal authority, I incline to think, without worked-out conviction, that the fact is what Fries intimates.²

I cannot off-hand say, whether in large genera the extremes differ more than they do in small genera; but it is likely they do so to some extent.

First, suppose the extremes about equally different or distant in both. It would seem to follow as a logical necessity, that the species are closer in the large genera: diag 1 2 3 4 5 6 7 8 9 Many sps.

1 - 3 - 5 - 7 - 9

1 - - - 5 - - - 9 Few sps.

1 - - - - - - - 9ramme

But, secondly, suppose the extremes different, on an average, in proportion to the number of species (which I do not believe); & that a representation would be thus:—³

1 2 3 4 5 6 7 8 9 Many sps

1 2 3 4 5 6 7 - -

1 2 3 4 - - - - - few sps

1 2 - - - - - - - The question will arise, on this supposition, Is this approximation of extremes real or conventional?—in nature, or in science only?

Whatever it may be in Zoology, I cannot think that in botany the groups are formed on any uniform principle; either orders or genera. The so-called ‘natural system’ in botany is to a great extent artificial or arbitrary; both orders and genera being (so to write) capriciously joined or separated. Hence, when we compare the number of genera in different orders, or the number of species in different genera, we are not truly comparing the facts of nature, but the capriciously technical arrangements of those facts by man. In some instances, Several very similar species are grouped together under one generic name. In other instances, very dissimilar species are grouped under one generic name. Is this natural, or is it only technical? Surely the latter! If so, what is the value of statistics founded on numbers in these groups? They are exponents of human science, rather than of nature’s facts or laws.

Still, as technical arrangements are intended to represent the resemblances among objects in nature, & doubtless do so in a considerable degree, altho’ not with minute exactness,—a large floral field, such as Europe or the American States, should afford tolerably good data towards your quest. In a botanical list so small as that for Britain, where genera are very unequally represented by the numbers of their species which are native here, statistical results are little to be relied upon except as mere indications,—something between possibilities & probabilities.

On the P.S. of your letter I may have some little to say another day. Saxifraga Andrewsii, a newly found or made species, may be a very striking instance of your apparent rule.—⁴

Sincerely Yours | Hewett C^l. Watson To | C. Darwin Esq