My dear Darwin

I was at Mr Smiths of Combe Hurst¹ last Sunday & met your relative Mr Mackintosh of W. Indies, who I liked very much.²

I there went deep into your Mss on variable species in big & small genera & tabulated Bentham after a fashion—but not very carefully—³ After very full deliberation I cordially concur in your view & accept it with all its consequences. Benthams book confirms you, though with modifications The larger genera I believe to be groups of more presently variable beings than the small & I think you have quite made good your point. Still I would not abandon the arguments against—for I still think that the disposition or rather the necessity of making more book varieties in large genera than in small is a very important fact.

I have also well considered Bentham’s exceptional orders, & am inclined to attribute that also partly to his idiosyncrasy—⁴ upon thinking well over his method of working, I have often seen that he will make rather hastily a new species in a large genus, of which a vast number of good species have recently turned up— The mental process is—“such & such a country teems with As-tragalus or Pedicularis—(which he has himself first elaborated)⁵ here is a new province of that country just supplied us with a lot of specimens & the chances are that heaps of them are new, & that more specimens will rather tend to prove doubtful new species to be distinct than the contrary”— It is not easy to explain to you how fully I appreciate this tendency in another person— but I am convinced it is so, & that that is the key to the Benthamian exceptional Orders—⁶

This does not however apply to Weddells Urticeæ which I must tabulate more carefully⁷ This was the case when Bentham & I did the Affghanistan & Tibet Astragali & Pediculariæ—⁸ he pronounced many new which I thought varieties.—always saying—Oh that country is the headquarters of Astragali—you must expect heaps of novelty.

In some passages of your mss you rather underrate I think the influence of associations of this & other sorts on descriptive systematists.— Look how Babington is eternally changing his mind—or rather his opinion (for mind he has none) as to genera & species—& it is so with all of us. Brown’s Specific work on Arctic plants is execrably bad & narrow-minded & I do not doubt attributable to some passing influence of this kind.⁹

On rereading your mss, I find the same objections as before, viz that you overrate the extent of my opposition to your method. My great desire was to put every possible objection as strongly as I could. I did not feel myself a dissenter from or opponent to your views, so much as a non consenter to them in the present state of my knowledge, nor till you had weighed my objections, wh. I thought of greater weight than I now do.

I am still dubious about the force of your argument against large areas & whole Nat Orders, or rather, of your argument in favor of small Floras, except you confine the latter to the plants strictly belonging to that area & exclude stragglers.¹⁰ Thus Panicum in Bentham ranks as a small genus & not variable, but those very species are the most variable plants in the world, elsewhere. I am however slow at taking up arguments of that sort, one way or the other.

I have still one objection to discuss—if as many middling sized genera are decreasing as are increasing how do we know which are which.— Do the species of big genera run to varieties in decreasing as in increasing. perhaps they should do so—yet you say “it is idle to discuss this”!

July 15^th. The E.I.C Examinations are cutting my time to shreds wh must account for some of the incoherency of the foregoing¹¹

I have had more time for thinking over the subject at odd half hours & have endeavoured to grapple with the whole question. That point of the hypothetical behavior of large genera when on the decrease puzzles me.

As a corollary to your Law, Large Natural Orders should have fewer genera in proportion to species than small; i.e. fewer defineable groups— Cruciferæ, Compositae, Umbelliferæ & Grasses bear you out in this—true no end of genera are made in them but they are bad— Other natural Orders are opposed.

I think I have thought of a better reason than you give for whole N. O^s being worse for your purpose than local Floras.— viz—1. that conditions do not go on varying with the area beyond a certain point:—there are limits to the combinations of Climate & soil A genus inhabiting 1000 square miles will survive such & such conditions, & under these influence form x species—all these conditions may occur in 100 miles of the said area, & adding the other 900 miles adds no more conditions. 2. Many large genera are absolutely confined to the Tropics or to temperate regions or to districts & do not stand in the same relation to one another as the mundane genera do— This I think you have expressed, but not clearly, any more than I have!

This would lead me to suggest the propriety of working one or two of your Floras by purging them of stragglers, & such plants generally as are typical of other climates & exceptional in this—of stragglers in short eg. Panicum— I think this process would intensify your results

I send the proofs from Linnæan Soc^y— Make any alterations you please—¹² if you want 100 or 150 pages of Linn. Journal for your abstract¹³ I think you should have it & we could publish it as a separate supplemental number of the