Think of your own case of Abyssinia and Fernando Po, and South Africa, and of your Lebanon case (354/2.

See "Origin," Edition VI., page 337.); grant that there are highlands to favour migration, but surely the lowlands must have been somewhat cooled.

What a splendid new and original evidence and case is that of Greenland: I cannot see how, even by granting bridges of continuous land, one can understand the existing flora.

I should think from the state of Scotland and America, and from isothermals, that during the coldest part of Glacial period, Greenland must have been quite depopulated.

Like a dog to his vomit, I cannot help going back and leaning to accidental means of transport by ice and currents.