Secondly, the same powers of flight which enabled the first bats to reach Madeira or the Canaries, would bring others from time to time from the African continent, which, mixing with the first emigrants and crossing with them, would check the formation of new races, or keep them true to the old types, as is found to be actually the case with the birds of Madeira and the Bermudas.
This would happen the more surely, if, as Mr. Darwin has endeavoured to prove, the offspring of races slightly varying are usually more vigorous than the progeny of parents of the same race, and would be more prolific, therefore, than the insular stock which had been for a long time breeding in and in.
The same cause would tend in a still more decided manner to prevent the seals from diverging into new races or 'incipient species,' because they range freely over the wide ocean, and, may therefore have continual intercourse with all other individuals of their species.
Thirdly, as to peculiar species, and even genera of bats in islands, we are perhaps too little acquainted at present with all the species and genera of the neighbouring continents to be able to affirm, with any degree of confidence, that the forms supposed to be peculiar do not exist elsewhere: those of the Canaries in Africa, for example. But what is still more important, we must bear in mind how many species and genera of post-pliocene mammalia have everywhere become extinct by causes independent of Man. It is always possible, therefore, that some types of cheiroptera, originally derived from the main land, have survived in islands, although they have gradually died out on the continents from whence they came; so that it would be rash to infer that there has been time for the creation, whether by variation or other agency, of new species or genera in the islands in question.
As to the rodents and cheiroptera of Australia, we are as
