It is self-evident that variation is constantly taking place, - sometimes for the better, sometimes for the worse, and that changes sometimes appear which do not affect intrinsic value. Most of the causes which produce variation are known, but the causes may not produce immediate noticeable changes. It may take several generations before the accumulated effects become marked. The careless breeder does not note these causes and their effects until degeneracy has gone so far as to make it advisable to introduce new blood, rather than make the attempt to correct the mistake by using the animals which have been varying for the worse.

It is not easy to determine by the subsequent effects the exact causes of slight variations or the exact time when they are introduced. There is one safe rule - endeavor to make all of the conditions of the animal superior to those which formerly prevailed; then when variations appear they are likely to be for the better. The reverse of this is equally true. Inferior conditions produce variations for the worse.

Keeping these facts in mind, the breeder is ready to begin his selection; provided, however, he has good judgment and a clear-cut mental photograph of the animal he seeks to produce. Selecting an animal does not improve it. The breeder simply takes advantage of the variations for the better due to causes which antedate his choice. By selection, he seeks to eliminate the influence of those individuals which have varied for the worse or have not risen, on the whole, above the average. By selection, he may progress faster than he would if he attempted to raise the progony of all the herd to a higher average level without selection.

A herd of bison, if changed from bad conditions and scanty pasturage to improved conditions and abundant food, will, in the course of a few generations, vary for the better; although natural elimination is likely to be more active under bad than under good conditions. It is, then, not natural selection, but improved food and environment, that have produced the variations for the better.

The farmer is usually advised to purchase the best male he can find and then practice rigid selection, if he desires to improve his herd or flock. There is either too much or too little in this brief recipe. The stream cannot rise above its fountain-head. While the introduction of a superior male into the herd may improve it for a short time, eventually the average of the herd will represent the powers for growth, development and production which are found in the food, coupled with environment, use and abuse. More than this, - the herd, in time, may actually fall below its environment and food-supply, because of the introduction of the improved blood; since it may take several generations before the improved blood adapts itself to conditions less congenial than were those in which it was produced. In earlier years, many a farmer did introduce an improved male into his herd, but did not improve the food and environment, and after a time found that no permanent advance had been made. Not seeing the cause of his failure, he condemned the breed to which the male introduced belonged, by saying that it was a pampered, tender breed for which he had no use.

If the farmers who desired to improve their herds had been told to select a male superior to the animals already possessed, and to mingle his blood with selected females, simultaneously improving food, care and environment, and then to select from the offspring, permanent betterment would have been secured. After having learned something in the school of experience of the methods which must be followed to secure improvement, the breeder would naturally seek to make use of a still better and presumably a higher-priced male, since he had learned to progressively and logically unite better blood, better food, better quarters, kindness and selection, harmoniously, giving to each factor its appropriate place and value. Evolution does not go forward by leaps and bounds, but by slow, almost imperceptible changes from generation to generation. The breeder of domestic animals would do well to follow Nature's modes of action.

So far, the selection from the offspring has been discussed, but of quite as much importance is the selection of the parents through which, in part, the betterment of the offspring is to be secured. If it is desired to secure offspring which when mature will reach sixteen hands, the result would be reached far more certainly by breeding a *female of fifteen and a half hands to a male sixteen and one-fourth hands high, than by breeding a female of fourteen hands to a male seventeen hands high. The progeny of animals radically different is seldom satisfactory. It is better to take two or three generations in which to attain the standard than to attempt to reach it at once by breeding very small and very large animals together. At the beginning, judicious selection is of prime importance. The one great mistake in horse-breeding is that of breeding large stallions to small mares; that is, little pains is taken to select such females as give promise of uniting harmoniously in the offspring with the blood of the male to which they are bred. It all comes to this, - that moderate variations through generations produce qualities which are likely to be permanent and potent.