The Ancon and Mauchamp races are only propagated by human industry. But these sudden deviations from a given type can also extend and multiply their numbers by themselves. It is well-known that South American oxen are descended from a horned Spanish race. Now, in 1770, a hornless ox was produced in Paraguay. In several years, according to d’Azara, this exceptional form had, as it were, invaded several provinces. Nevertheless, the race is far from being in favour, because the absence of horns renders it less liable to be caught by the lasso, so that its destruction was attempted. It was, therefore, evidently propagated spontaneously. Whoever has the smallest acquaintance with embryogenesis will have no difficulty in understanding that conditions of life act especially upon organisms in respect to their formation and evolution. However, their influence upon an animal, even when full-grown, is sometimes quite as marked. Our sheep, when transferred to America, generally become acclimatised without undergoing great changes. Their fleece, particularly, is retained. But in the plains of the Meta it is only retained on condition of the sheep being regularly shorn. If they are left to themselves, the wool becomes of a felty nature, is detached in flakes, and is replaced by a short, stiff, and shining hair. Under the influence of this burning climate, the same individual becomes in turn a woolly and a hairy animal. Now, innateness, as Prosper Lucas conceives it, cannot be appealed to in the case of changes undergone by a full-grown animal, whilst the action of conditions of life is here incontestable. VI. We have just pointed out how heredity and conditions of life can give rise to a _variety_. Now, the individual which has commenced to deviate from its original type becomes in its turn a _parent_; it tends to transmit to its offspring the exceptional characters which distinguish it. The same facts are repeated in its offspring; and, at each generation, the actions of the conditions of life are added to each other. Thus every time heredity transmits the sum of these actions to the following generation. The faintest modification increased from father to son sometimes leads to most marked changes. Our European oxen, in the hot plains of Mariquita and Neyba gradually lost their hair, at first became _pelones_, and would soon have formed an entirely naked race, if the _calongos_ had not been regularly killed. Again, pigs which have become wild in the Paramos have acquired a kind of wool under the action of a continuous, but not excessive cold. The Guinea dog and the Esquimaux dog present an analogous contrast between races of the same species.