THE abundance of birds on the four largest islands of the West Indian archipelago, where indigenous mammals are almost limited to rodents and bats, has often suggested the conjecture that the ancestors of those islanders must have been immigrants from the east coasts of the American mainland; and that theory seems to be confirmed by two facts: the identity, or similarity, of numerous Mexican and West Indian species, and the circumstance that those analogies include so many swift-winged birds.
There are no woodpeckers in the forests of the Antilles, and only two species of large gallinaceous birds, but a prodigious variety of
pigeons, swallows, finches, and crows. The alcedos (kingfishers) are scarce, but the blackbirds so numerous that some of the countless species seem to claim a South American and even transatlantic ancestry. The restless estornino of the Cuban highland forests, for instance, might be mistaken for a varnished starling, resembling the Sturnus vulgaris of western Europe in everything but the more brilliant luster of its plumage. The curious codornilla, or dove quail, too, has its nearest relatives on the other side of the Atlantic, in Syria, Arabia, and the foothills of the Atlas. It builds its nest on the ground and, judging from its appearance, would seem to form a connecting link between the doves and small gallinæ; but its wings are those of a pigeon, and with the assistance of a, northeast gale may possibly have carried it across the ocean.
In studying the geographical distribution of animals, we may estimate the prevalence of special genera by the number of their varieties, or by the aggregate sum of individuals, and in the latter sense the migratory pigeons of our forest States once nearly outnumbered all the other birds of North America, though the family is limited to five or six species. But in the West Indies the Columbidæ predominate in both respects. Cuba is a country of wild pigeons as preeminently as South Africa is a land of pachyderms and Madagascar