Dear Darwin

I will only continue to correspond on the conditions that you will write as often as you want & wish & think I can be of the smallest service— your questions are most profitable for me to work. I saw Bentham today— you are half right & half wrong in your surmise—¹ thus Bentham has omitted all very large & anomalous genera that for size may be considered equivalents of Nat. Ords. as well as being Nat Ords from peculiarity—as Begoniaceæ for instance, a bona fide nat: ord. with perhaps 160 species, but which neither Bentham nor I would make more than one genus of.— Genera & nat ords are however to such an extended degree arbitrary as to their limits, that these questions are all most difficult to work. There is another light however in which you must view these catalogue—which is this, that except you include all the anomalous genera in the Veg. Kingd:² it is not fair to take the greatest; because, in taking a miscellaneous selection, the great genera like Begonia Nepenthes & Cassytha will always come up to mind from their being so familiar, whereas the little known anomalous genera of few or one species are not recollected & many of them not known to even a learned Botanist. In other words for every anomalous genus big with species there are 50 of one or very few species, but a person in making a miscellaneous selection jumps at the big ones & overlooks the small.

My grave objection to the whole process is, that if you multiply the anomalous species by 100 & divide the normal by the same you will then reverse the names, for anomaly is to a great extent synonymous with rarity or representation by few species. EG. as plants go now adays, Lepidodendron is an anomaly; but as years ago i.e. in the coal epoch no doubt a Rose would have been an anomaly— also substituting space for time & we have Proteaceæ anomalies in the S. American Flora & Fuchsias in the Australian— Destroy Australia & the Cape & what a deuced odd anomalous order Proteaceæ become.— perhaps however I am (to you) reasoning in a circle, & arguing from the conclusion you want to arrive at. We must take the world & its Veg. Kingd. as it is however, & I can see that a great deal should be made out of a proper tabulation of anomalous genera.— The curse of Botany is the extreme difficulty of determining the affinities of plants & of limiting groups— they represent one another & inosculate & repeat forms & present an impenetrable barrier to the distinguishing between affinity & analogy. The three great groups Acot. Monocot & Dicot. are absolutely defined & limited; but amongst Dicots there are but few absolutely limited natural orders, & not only do most of the orders blend in every way but the most distinct contain species & groups which imitate in outward appearance plants with which the order to which they belong has no real affinity this enhances the difficulty tenfold. In all my experience I know no science so difficult as that of the affinities of plants.— it can only be got by dissection & study; it cannot be learned & what is worse cannot be expressed except by details.— Upon the whole I am inclined to take Benthams list as a fair selection but the only other way I can suggest of working out the point is to take single orders & seek the proportions that the anomalous genera in them bear to the typical genera.

To return—you will therefore understand that Bentham did not exclude large genera simply because they were large, but he did so where these were large & also represented Nat. Ords. I quite agree that your plan is to take anomalous forms of whatever value, but if you only take a selection from the Veg. K. you should beware of including the very big ones or you will have an undue preponderance of them.

With regard to splitting Australia,³ you say I underrate the consequent effect on all the species— now I rather shirked that as being a purely hypothetical point; the fact that three fourths? would be killed & that a greater proportion of those common to both would be killed than of those peculiar to each, appear to be the great consequences— do you mean that besides this some species now distinct would assimilate, or that those now the same would be altered in each & in different ways?

I send you the results of the Aquilegia work,⁴ on which Thomson & I spent many hours & he good three weeks of indomitable perseverance with every accessable book & specimen wild & garden dried & alive It is badly composed, but was one of the first of Thomsons trials at writing out his results: three revises were corrected after this but I am sorry we have not a clearer one: therefore excuse errors in this & return it at your leisure. it is not yet published— we shall send you a copy when it is. there are about 400 more pages printed of the Flora Indica; I am as sure as can be that there is but one European & Himalayan Aquilegia out of which 30 are made & I suspect that Poeonia & very many other genera are