ach successive stage of modification and
improvement, all the individuals of each variety will have descended
from a single parent. But in the majority of cases, namely, with
all organisms which habitually unite for each birth, or which often
intercross, I believe that during the slow process of modification
the individuals of the species will have been kept nearly uniform by
intercrossing; so that many individuals will have gone on simultaneously
changing, and the whole amount of modification will not have been due,
at each stage, to descent from a single parent. To illustrate what I
mean: our English racehorses differ slightly from the horses of every
other breed; but they do not owe their difference and superiority to
descent from any single pair, but to continued care in selecting and
training many individuals during many generations.

Before discussing the three classes of facts, which I have selected as
presenting the greatest amount of difficulty on the theory of "single
centres of creation," I must say a few words on the means of dispersal.

MEANS OF DISPERSAL.

Sir C. Lyell and other authors have ably treated this subject. I can
give here only the briefest abstract of the more important facts. Change
of climate must have had a powerful influence on migration: a region
when its climate was different may have been a high road for migration,
but now be impassable; I shall, however, presently have to discuss this
branch of the subject in some detail. Changes of level in the land must
also have been highly influential: a narrow isthmus now separates two
marine faunas; submerge it, or let it formerly have been submerged, and
the two faunas will now blend or may formerly have blended: where the
sea now extends, land may at a former period have connected islands or
possibly even continents together, and thus have allowed terrestrial
productions to pass from one to the other. No geologist will dispute
that great mutations of level have occurred within the period of
existing organisms. Edward Forbes insisted that all the islands in the
Atlantic must recently have been connected with Europe or Africa, and
Europe likewise with America. Other authors have thus hypothetically
bridged over every ocean, and have united almost every island to some
mainland. If indeed the arguments used by Forbes are to be trusted,
it must be admitted that scarcely a single island exists which has not
recently been united to some continent. Th