From Charles Lyell¹   28 August 1860

The grand argument from absence of mammalia & batrachians in Oceanic islands is probably felt to be strong by Owen as he has not ventured to impugn it and therefore declined to bring it into notice.²

It is a very telling argument & I think just & true, tho‘ if I had been helping Owen & the Bishop I fancy I could have suggested objections but I have not thought them out yet, & have not therefore troubled you with them, thinking I sh^d be able to answer them myself.

I remember however it struck me that the putting it even as possible that the Galapagos or Cape de Verds could have been joined to the nearest Continent was dangerous, as in the case (p. 390) they must have retained some of the smaller mammalia. Also if atolls be remnants of sunk continents would not some of the old continental land quadrupeds have survived or have been slowly modified into amphibious & volant species.³

Also, if as I believe the Canaries & Madeiras have been islands ever since the Miocene period (& the Azores), would not the theory of Nat. Selection have time to modify any vertebrate animal into the lowest of the Lyencephala in such a lapse of ages.⁴ Or if not, does it not give us a sort of scale for the vast geological time required for such a transmutation. No Galapagos saurian, having risen to a mammal. Perhaps you will say that this would be a greater puzzle to the out-&-out progressionists who may be converted by you to transmutation. Bronn,⁵ e.g., for you do not believe in that constant advance, as an ever working tendency, & allow for stationary & retrograde movements.

You would prefer I conceive to derive Eocene mammalia from the Microlestes of the Trias rather from any reptilian of however high a grade e.g. a deinosaurian. You w^d rather conceive a bird to be turned into a mammal than derive the latter from a reptile. The Microlestes, or its nearest living analogues (marsupials) w^d. in its embryology go through the bird-state or likeness before it became even a non-placental mammifer?

But if an island be very ancient the bats would be modified into mammalia of other genera & orders long before any bird could be turned into a mammal. Here comes the question, how old is the oldest isolated Oceanic island? The miocene littoral shells of Grand Canary & of Porto Santo give a miocene date to those lands, already in all probability separated from the Continent as now. But there are no bats there.

New Zealand is so large it probably existed from Eocene times as land?? Why did not bats getting there diverge into a dominant mammalian volant fauna & some lose their wings & become non-volant mammalia? Or why did not seals & marine mammalia turn terrestrial in their habits if they had such a geological period as the post-miocene for nat^l. selection to work in. The seals swarmed in Madeira before Man drove them away. This difficulty strikes me as greater in as much as the rate of change in mammalia is more rapid than in inferior grades & there have been in Europe several changes of mammalia since the Upper Miocene period. The adaptation of a seal which is already half terrestrial, requires less time than the conversion of Hearn’s bear into a whale which they are never tired of quizzing you for.⁶ I shall be glad if you will tell me whether these objections are of any value. Probably you have not thought the evidence of the antiquity of any islands good for much. Yet it seems necessary to grant a good deal of time to cause so many endemic shells in Madeira & Porto Santo & land shells would change much more slowly than mammalia. Still I grant that the conversion of a phoca into a terrestrial quadruped is very different from one helix into another, but ought not the seals to have begun in ancient islands to keep somewhat longer out of water. These are perhaps unfair, imaginary difficulties & ought not to weigh against an hypothesis which explains so much as does transmutation in the case of absence of mammalia & frogs from remote islands while they swarm in islands which we know have been united in newer pliocene times.

If there was time in Galapagos to convert a marine Amblyrhincus into a land species or vice versa. why not a phoca into a land species or other cetaceae into terrestrial. Did the land Amblyrhincus come first into being & of what S. American Saurian was he or his aquatic congener a modification? If all these questions apply equally against independent acts of Creation they will not be very damaging. But when the free agency of the First Cause is allowed to come into play, it may be said that no mammalia were wanted because it was foreseen that Man would come with his attendant train of useful mammals, taking even the camel with success into the Canaries.