Assuming another case, and this time of a bitch with yellow eyes. Now the average breeder would select a dog that had eyes of good color, without considering for a moment that that dog may be an "accident," - or, in other words, from a yellow-eyed family, - but he who has had experience and profited by it would satisfy himself on this point, and if the dog came from such a family he would choose one from a family good in eye.

That there may be no mistaking the meaning of all this still another case is assumed, and of a pointer bitch light in head. Her owner, an intelligent breeder, instead of choosing as a sire a heavy-headed dog mates her to one that is perhaps a little on the light side, and the puppies come right. Why did he do this? He knew what there was in his bitch's pedigree. In a word, he knew that she was closely related to a strong-headed family, and that the sire he had selected for her was also from a family that were good in head. But had he selected a short-faced and thick-headed dog of a short-faced and thickheaded family, the result would doubtless have been far from his liking.

The obvious conclusion is, that in selecting a sire one must not only be familiar with the available dogs themselves but have a good knowledge of their family history; moreover, he should know whether or not their offspring very generally resembled them or some of their ancestors. Some sires and dams, in fact, seem to have but little individuality, as shown by their inability to reproduce themselves. Two inmates of the writer's kennels plainly testified to this fact. Both were mastiffs - dog and bitch two removes from "Crown Prince," and notwithstanding they had dense black muzzles, no matter how mated they almost invariably "threw puppies" having the same peculiar chocolate-colored muzzle of their noted ancestor.

Manifestly, therefore, a good, shapely and well-marked dog may prove but an indifferent stock-getter, while, on the other hand, one with a glaring defect - as bad color of muzzle, size of ear, expression of eye, etc. - may turn out admirably. And although such results, good or bad, may sometimes be purely accidental, as a rule, to which there are not many exceptions, they are due to that peculiarity of "throwing back."

Experience has shown that admirable qualities must exist for many generations in order to render their perpetuation highly probable. For instance, if a dog has a grand head yet his sire and grand-sire were deficient in this quality, then the chances are against his offspring being endowed like himself; and especially narrow are they if their dam is weak in head. Were she so and only two dogs available, one with a grand head yet of small-headed ancestors, and the other small-headed but of a family noted for good heads in the two generations before him, her owner would probably have the best results from the use of the latter.

Another product of experience, is that to breed out some defects is much more difficult than others; and as a rule, the resistance is in proportion to their duration in the family. In other words, a fault of two generations is much less easily eradicated than a fault of but one.

Here it is well to emphasize the fact that weak points are as likely to be transmitted as strong ones; also that structural defects, as deformities of limb and other vices which result from errors in management during the growing stage, are liable to be passed on. And, in truth, it is oftentimes so difficult to breed out acquired abnormal peculiarities one would be seldom justified in choosing a sire that had any such that were very serious.

In breeding non-sporting dogs the form and proportions are the main objects of study, and efforts are generally centred upon their improvement alone, but with dogs to be used for special purposes, as in the field, there are more characters to be developed and therefore a greater number of difficulties to contend with. Consequently, oftener in this line than in the other breeders meet with incompati-bles, and in many instances they must fix upon a few special qualities, endeavor to bring them to a high degree of excellence, and in the mean while practically neglect others. And in doing so they frequently find it necessary to take risks; that is, in order to improve one quality they are often obliged to put another in jeopardy; but he who is wise never does this unless the gain promises to be much greater than the loss.

Setter breeders well illustrate this principle of "give and take." They have produced in what is known in this country as the Llewellin setter great speed and dash, yet have sacrificed much of the beautiful setter quality that is seen in the Laveracks.

But in breeding for the field, while it is desirable that both sires and dams be well broken, to base selections purely on excellence as workers is clearly wrong if real improvement is the end in sight, for, manifestly, there is no chance for form if one breeds for nose, another for speed, and so on, ignoring the points of the breed. And the reader should unhesitatingly accept that he can never justly sacrifice type for any educational quality. Moreover, he should be duly mindful of the fact that less importance is to be attached to a killing pace and more to other essential qualities, also, that he cannot expect greyhound speed in setter form, and if he is to have one he must give up the other.

Every beginner should start out with a fixed purpose of breeding, not for market merely, but to advance his special variety of dogs to a higher grade of excellence. Even if he is so fortunate as to be able to indulge his inclinations to their full extent, he should not invest largely at first. Three or four brood bitches will be quite as many as he can do justice to while yet inexperienced, and to him they will prove more profitable than would twice this number.

If he has a dog of his own he will likely use him in every instance, and if he does so he will be liable to go backward as often as forward, no matter how good the same or how good his mates. In fact, to buy a dog and breed him to every bitch in the kennel is one of the worst mistakes which a breeder can make, and one which, as a rule, destroys many of the chances of eminent success.