I found yr letter here & thank you for intending to send me Asa Gray’s last article which I will return without delay as also the M.S. on dogs.² Agassiz in some one of his Essays leans to Pallas’ views,³ which always struck me as being rash in him who is for the sanctity of species, whereas if any thing could show Nature’s entire indifference in regard to the violation of such supposed sanctity it would be that those now existing should be of multiple origin.

You need not be afraid of my starting any theory of successive creation of types. I have never had any leaning towards a compromise and thought Carpenter very inconsistent in going so very far with you, & then concluding that there was probably in the case of Man, an exception, whereas there would be no end of all the morphol^l. & embryological arguments if such a concession was made.⁴

But what I asked or intended to ask in my last letter was different—perhaps you meant to reply that you had no fixed opinion on the subject—but I wish to be sure.— Can we assume as at all probable that all mammalia came from one original Stock instead of several distinct mammalian-types each developed by small & successive modifications out of lower ornithic reptilian or perhaps monotrematous prototypes. It would greatly simplify matters if single & exclusive areas could be assigned or even speculated upon, on some even slight data, or if single periods could be proposed as those of the first coming in of mammalia. But as I understand your views this is not very probable. If for example we found before the triassic microlestes, some monotremes in Permian or Carbonif^s. beds & fixed on them as the probable starting point of mammalia, what influence would the development of a mammal in Asia or Africa have in preventing some other line of monotremes from improving into a totally different genus of mammal in Australia or America & supposing the mammal first formed, say in Asia, to be extinguished, what is there in your system to cause us to despair of the higher or placental grade from ever being evolved because after a geological period, the earlier formed mammalia died out.

I have always expected to find mammals intermediate between Stonesfield & Eocene, but I also looked for them before Stonesfield. When therefore they turned up both in Trias & in Purbeck, it agreed with my anticipations.⁵ The Triassic surprised some people more than the Purbeck, for they held the doctrine, propounded by you in p. 316, once begun the group is continued till it dies out.

I wish you could give the slightest reason why it should not begin more than once in more than one place. I incline to think it has not, but why? according to the principle of selection why when once in any one quarter of the globe at any one period the step in advance has been taken are the inferior types elsewhere to be checked, & not to presume to work up into any genus of corresponding grade & class? A frog has a vast deal of one type in him even the double occipital condyle, to say nothing of his limbs & digits & the absence of a tail, tho‘ he is low in the reptilian scale, (but it is true, he does not as yet go far back in time) but take some triassic reptile or bird I cannot see why they are never to be modified into something higher because a microlestes has entered on the stage.⁶ Two independent chains of development would not be discordant with your machinery of selection?

I am not aware of any fossil bats or rodents in Australian caves, so the small antiquity of rodentia may be a reason for their not having given origin as yet to Gyrencephala & does there not seem some connection between the low grade of Lyssencephala as the only occupants of that vast Australian continent?⁷

Can you not refer me to some book or paper where I can find how many plants may have been extirpated in St Helena & may I not presume that as many or more endemic insects have been annihilated there besides some land-shells?

As to Australia, I cannot help thinking that the Marsupials must in the long run (if Man were away) be best fitted for that continent & that after European or Asiatic Gyrencephala had been introduced & then left to themselves, they w^d. during some 4 years drought, after a century or much more, have all perished & given way before their pouched rivals, who as Owen has somewhere suggested, w^d have migrated with their young in the pouch, wholly indifferent to the long leaps of the mother.⁸ This w^d not apply to bats, but perhaps you may be able to object that some of the native placental rodents w^d be no better off than a lion or an Elephant in a 4 years drought?

In y^r chapter on grafting you might, it struck me, have alluded to those wonderful cases of nat^l. grafting recorded in the fir-woods of Germany. I think I could find a reference to it, where a totally leafless stump keeps alive by its roots, borrowing from the roots of adjoining fir-trees till the bark closes over.⁹ It answers the question why Nature has given such a power & was it for the mere sake of human gardeners.

I am not going to press the objection about the non-variation, adaptation & advance of bats & rodents, so as to destroy the weight of y^r. argument respecting absence of other mammalia in islands, as I cannot conceive such absence not being due to the want of migratory powers of mammalia, & that alone w^d not do, unless the coming in of new mammals was by inheritance from pre-existing types. But it is better to show that one does not overlook objections.

I think I asked you once before whether the admission of the multiple origin of dogs w^d. not impair one’s faith in the single origin of the human race, & as you did not reply, you perhaps meant that you had nothing to hazard in the way of speculation. You remember a passage in Agassiz about negros being black where the Chimpanzee is black, & yellow in Borneo? where the orang is yellow, or something to that effect.¹⁰ I could hunt out the passage if you do not remember it. As he does not connect Man genealogically with any anthropomorph, he says this without any hesitation.

To shorten correspondence I will imagine the best answer I can for you & beg you to try & strengthen it.

It is so many millions to one against the right variation offering itself in the right place & time that many geol^l periods must pass away before the first mammal is formed & as an equal lapse of ages w^d occur before a 2^d. offered, the first will in the meantime have spread over the globe & branched out into many genera, w^h. being of high grade w^d. keep down any incipient mammals, just as homo sapiens when once evolved w^d. take good care that no other anthropomorph sh^d improve into rationalism & compete with him for the good things of this world. No island has existed for so long a time as that which separates the Lower Miocene from the human epoch, therefore the chances have been far too many against transmutation of any creature of an inferior class into one of a higher class on an island, more especially as bats & rodents w^d help to check any signs of an invasion of the prerogatives of the dominant class, they being already there to represent the vertebrate aristocracy.