skip to content

Darwin Correspondence Project

From H. C. Watson   2 October 1855

Thames Ditton

Oct 2d. 55

My dear Sir

I had felt an inclination to remark on some of the points in the letter of 26 augt., with which you favored me. Visitors in the house for a month, & other circ\ {e}s, have come in the way; & you may now not fully recall what you wrote.

There is doubtless ample reason for your viewing “the existing Fauna & Flora as a mere fragment” of the Faunas & Floras of all time geologically. But it seems to me rather a strained view of this received fact, to apply it as a reason for taking the numerical data of the British Flora,—a fragment of the existing Flora,—to be adequate to your object of ascertaining whether species are closer in large genera, than they are in small genera. Some of the small genera in Britain are large genera in Europe or the Earth. By looking to Britain only, these genera would be entered in the wrong category, & appear to evidence the opposite of what they do evidence.1 It may have been, indeed, that present small genera were once large genera, & that existing large genera were contemporaneously small genera;—change of time doing on the grand scale, what change of place or country now does locally.

I think that the extreme species of large genera do differ more than the extreme species of small genera; so that, in reverting to the numeral illustration, a large & a small genus might be represented thus: diag 1 2 3 4 5 6 7 8 9

3 4 5 6 7ramme

But it seems highly probable that the difference or distance between the extreme species is not proportional to the number of species included in a genus; the species of the large genus being often closer. Say, thus:— diag 1 2 3 4 5 6 7 8 9

2 - 4 - 6 - 8ramme

In many small genera the species may closely resemble each other, it is true; but in many large genera they are so very close as to almost defy botanical attempts at distinctive characters. The difficulty is doubtless much increased by the fact that, in a genus of 50 species, each one has to be distinguished from 49 others; while in a genus of three species, each one has to be distinguished from only two others. Still, I conceive that some part of the difficulty would be properly attributed to the greater closeness of the species.2

But what is a close species ? Those which Botanist A would call the closest species, Botanist B would deny to be species at all; deeming them varieties. Those which Botanist A would deem species of only secondary closeness, would by Botanist B be pronounced the closest species he could admit. On which of these Two would you rely?—Jordan, the “splitter”;—or Hooker, the “lumper”?3

There is a peculiarity in botanical classification which bears importantly on your question. To define or distinguish Orders and Species any part of the plant, any structure, &c., is taken into account. But in the intermediate step or group of Genus only the reproductive structures are used as a general rule. Plants very similar in their reproductive organs (or flowers & fruit) may & do differ greatly in their leaves, roots, stems, mode of growth, &c. Thus in Veronica, where there are some very close & even disputed species, there are others remarkably dissimilar. In Mentha, the range of dissimilarity is much less. In Papaver, looking only to the 4 British species, the similarity is very close & comprehensive, altho’ the species are readily distinguished by one or two slight tho’ constant characters. In Fumaria, the similarity is also very close & comprehensive, & the species barely distinguishable by technical characters.

Perhaps the 1st. Category or degree of closeness would be got by taking all those plants which are called Species by some botanists, varieties by other botanists. These may be species in part, and clearly can have no very salient differences.

A second category would be got by next taking those species, usually so deemed, which have some one or two pretty clear characters of difference, but agree in other respects. Thus: Scleranthus annuus and perennis, Drosera rotundifolia & longifolia, Chrysosplenium oppositifolium & alternifolium, Arenaria serpyllifolia & tenuifolia, are pairs of species generally much alike, yet readily distinguished by the one character expressed in the specific name.

3d. As a counter example, of general difference, read the description of two Bramble-allies:— diag 1. Rubus Idæus (Raspberry) — Stem biennial, woody, prickly,

many-flowered, tall. Leaves compound.

2. Rubus Chamæmorus (Cloudberry) — Stem annual, herb-like,

unarmed, single-flowered, short. Leaves simple.ramme

But such Categories really glide into each other when a whole Flora is looked at; some pairs of species having one character of distinction;—some, two;—some, three;—&c.

Sincerely Yours | Hewett Cl. Watson C. Darwin | Esq

CD annotations

crossed pencil
‘It is true that I do assume this; but there are some good arguments in Favour: Mem Forbes.’ added pencil; ‘I doubt whether I have argued out that large genera were once small. I knew Forbes’ added ink
2.7 the Earth.] ‘so small mundane genera, were probably once large’ interl pencil
scored pencil
2.12 locally.] ‘I shd expect same law wd. apply to a mere field, only here the space would not be enough to give average, as no one supposes that every large genus has close species.’ added pencil
6.2 species at all … varieties.] scored pencil; ‘This wd be a great difficulty, if species have real existence; but I look at vars & species as running together; so the facts given by a “lumper” would apply only to forms more unilike. I shd expect it would even apply to genera, if it cd be tested: & I suspect that it does hold in genera in [most] Families, but then so vague what a genus is.’ added pencil
‘I am astonished at this: would leaves & stems be used in order???’ added pencil; ‘It comes to this that the genera are not natural!’ added ink
double scored pencil
underl pencil
8.3 These may … differences.] double scored pencil CD note added to paragraph 7:4 The only bearing which I can attach to this, is that genera are not strictly founded on aggregate amount of difference: these 2 or more genera are founded on differences in reproductive system, whereas the real & total amount of difference in the species of these 2 or 3 genera, is not greater than in some single genera, which may include equal or greater diversity, but which are included in one genus, because reproductive organs are similar.— This comes to whole essence of natural classification.—why reproductive organs, are considered most important—which I fancy turn [‘out’ del] on permanence.— Granting that the classification is false, then my supposed law would have some mysterious relation to reproductive organs, viz then there wd. be most close species in the groups in which reproductive organs resembled each other.—


CD sought to justify his use of local floras in Natural selection, pp. 156–7.
CD cited Watson’s view and used a similar numerical illustration in Natural selection, p. 147.
Alexis Jordan and Joseph Dalton Hooker. See letter from H. C. Watson, 13 August 1855, n. 3.
This note is keyed to paragraph seven by the symbol ‘XAX’ on the note and in the text of the letter.


Expresses his general opinion on the relative closeness of species in large and small genera. Warns that the size of a genus is dependent upon the locality and extent of the flora studied, that definitions of close species are not consistent, and that peculiarities of botanical classification will influence any attempt to assess the comparative closeness of species in different genera.

Letter details

Letter no.
Hewett Cottrell Watson
Charles Robert Darwin
Sent from
Thames Ditton
Source of text
DAR 181: 30
Physical description
4pp †† CD notes, 1p

Please cite as

Darwin Correspondence Project, “Letter no. 1758,” accessed on 20 March 2019,

Also published in The Correspondence of Charles Darwin, vol. 5