As in the case of fowls just noted, we find that domestic ducks of the same species, after a separation of several thousand years, breed freely together. But our domestic ducks are not, like the fowls, all of the same species, and if individuals of different species are paired they produce only a few weak hybrids. Our domestic geese are probably descended from two wild varieties, but races that were not brought together for thousands of years after they were domesticated are perfectly fertile together, while when mated with the American Wild Goose, which is not domesticated but will breed in captivity, they produce only hybrids. The general resemblance between geese and ducks is very striking, yet they will not breed together at all.

A comparison of these facts indicates that while three thousand, or even five or six thousand, years of separation may not be enough to break down the natural affinity of varieties of the same species, separation and difference of development will eventually make of varieties distinct species, a union of which will produce only hybrids, while a longer separation and further increase of differences makes the break between the species absolute and they will not breed together at all.

Natural varieties. A species having developed as a variety of an earlier species will continue to develop as one variety or as several varieties, according to conditions. If a part of a species becomes so separated from the rest that intercourse ceases, each division of the species may become a well-defined variety.

Varieties in domestication. How a species when domesticated breaks up into varieties is well illustrated by the case of the fowl. The original wild species has long disappeared, but there is good reason to suppose that in size and color it was something between a Brown Pit Game and a Brown Leghorn. The birds were smaller than most fowls seen in this country to-day. The prevailing color was a dull brown, because that color best conceals a small land bird from its enemies. Fowls that were domesticated and given good care and an abundance of food would usually grow larger than the wild stock. Thus if any person, or the people generally in any community, systematically gave their fowls good care, a variety of unusual size would be developed.

Different colors would also appear in the flocks of fowls, because the birds of unusual colors would be protected and preserved, instead of being destroyed as they usually are in the wild state. Other peculiarities, too, such as large combs, crests, and feathered legs, would be developed in some lands and neglected in others. This is how it happened that after thousands of years in domestication the races of fowls in different parts of the world were quite different in size and form, but alike in being of many colors.

From a species in this condition modern poultry breeders have made hundreds of distinct varieties. The easiest method of making a variety in domestication is to select specimens for breeding as near the desired type as possible, and to breed only from a few individuals in each generation which come nearest to the ideal type. In this way a variety that breeds quite true to the type may be established in from three or four to eight or ten years, according to the number of characters to be established as distinctive of the variety. Varieties are also made by crossing unlike individuals. This process is longer than the other, and sometimes requires a series of crosses to produce specimens approximating the ideal sought. After such specimens have been obtained the method is the same as in the first case. A variety is commonly considered to be well established when the greater part of the specimens produced are easily identified as of that variety. But no domestic variety is ever established in the sense that a species is. All are artificial, produced by compulsory separation and preserved only as long as it is continued.

Classification of domestic varieties of birds. Domestic varieties of all kinds of live stock were at first mostly shape-varieties; that is, the individuals of a variety were alike in shape but of various colors. This is the case still with some varieties. These shape-varieties are mostly the common types of certain countries or districts. Thus the Leghorn fowl is the common fowl of Italy, and the Houdan is a type common in a small district in France. Such shape-varieties are called breeds. When other types were made by crossing such breeds they also were called breeds.

When people first began to be interested in the improvement of live stock, the popular idea of a breed was that it was a domestic species, and there are still many people who hold this view. This popular misconception of the nature of a breed is responsible for much of the inconsistency and confusion in the ordinary classifications of domestic varieties. To it also is due the use of the term "variety" to apply especially to color-varieties, which are the principal divisions of breeds.

In the classification of domestic birds a variety is properly a color-variety of a breed. Thus in the Plymouth Rock breed there are six color-varieties—barred, white, buff, partridge, silver-penciled, and ermine (called Columbian); and in Fantail Pigeons there are six color-varieties—white, blue, black, red, yellow, and silver. Birds of the same breed (shape) and the same variety (color) may differ in some other character, as the form of the comb or the presence or absence of feathers in certain places. In accordance with such differences varieties are divided into subvarieties. Thus, in Leghorn Fowls the brown, white, and buff varieties have single-combed and rose-combed subvarieties.