It would seem from the practice of the infinite majority of breeders that they, in the beginning, accepted as a fixed fact that "like produces like "with positive certainty, and satisfied that in this familiar aphorism was comprised the one and only essential principle, in the application of which none could stray, they were quite content and had no desire to go farther into the subject. It is not surprising, therefore, that but few have succeeded while many have failed, and that a large proportion of the best dogs have been what might without impropriety be called accidental creations.

Some breeders consider pedigree of first importance and mate accordingly, practically ignoring the question of suitability, also the fact that with good pedigrees on both sides the puppies are often very poor specimens. And these rely mainly on reversion; or in other words they confidently expect that even if the dogs chosen are not themselves all that is desirable they will yet be sure to "throw back," and that their offspring will resemble ancestors, more or less remote, which were good.

Other breeders believe that success at shows is a guaranty of all the most desirable qualities, therefore they invariably seek sires among prize-winners, without thought of fitness or questioning whether they won in good company or under competent judges. And with them there are but two accepted laws in breeding, namely, "like produces like" and "breed always from the best" - the "best" being those of visible merits without considerations as to the qualities of the ancestors or conformity of the individuals selected to the same general types of their families.

Members of both classes are occasionally successful, yet when so they scarcely deserve credit, for they have no ideals nor real systems, and always follow, never lead.

Of course their expectations have foundations, for it is evidently a law of animal organism that the offspring shall inherit characters of parents, but this does not mean that they shall inherit all the characters, nor even one or more of the most desirable, for there are other laws the influence of which may be predominant and for the time being at least obscure this hereditary tendency. Again, while the offspring may be said very generally to resemble the parents the resemblance is not, as so many assume, confined to the outward form and visible characters, but as often, doubtless, manifests itself beneath the surface, and without evidence except such as appears in psychical qualities.

"Breed to the best "is of course a golden rule, provided it is rendered rightly - that is, the selection is consistent and the breeder is influenced not alone by obvious excellence of the individual but by the family history as well. In other words, he has a proper conception of it who looks for a combination of qualities when seeking a sire, and considers not merely the dog himself and his dominant characters but the characters of his family, the constancy with which good qualities have been transmitted, whether the existing perfections will compensate for ancestral defects, etc.

He who would improve his dogs by developing their most valuable qualities and fortifying them with others, and so give evidence that he is something more than a breeder in name merely, must have a conception of the qualities that constitute perfection - an idea of what he wishes to create, the ideal form he would mould. He must also be able to detect slight variations in form and qualities within; moreover, have a knowledge of the fundamental laws of animal organization, and especially those that relate to inheritance.

With these and good judgment and perseverance he can feel confident of success. But their acquirement means systematic observation and an abundant material for study, consequently it is scarcely surprising that there are so few real breeders.

The one who possesses these eminent qualifications seeks the dog that is most likely to correct the faults of his bitch and at the same time preserve her good qualities in the offspring. But he does not follow the custom which seems so prevalent among breeders and choose-always a dog that is strong where his bitch is weak. For instance, if the average breeder has a bitch whose muzzle is too long his choice is a dog with a good muzzle; or if she is "leggy" he seeks a stocky dog with plenty of bone and muscle. Again, if breeding for color and his bitch is rather light, he chooses a dog that is inclined to be dark. And in all his selections he considers merely individual excellence; with the result that only comparatively rarely does he breed puppies nearly as good as their parents.

But the knowing breeder does not pursue this course, tor he goes back of the individuals and is influenced by the characters of their families; and while he prefers a dog that is not only good himself but came of stock that was invariably good, he values family excellence above that of the individual. In other words, if his bitch has too long a muzzle, his choice is a dog that comes from a family that were good in muzzle, rather than a dog which exhibits this quality but is of a family that were poor in muzzle.

He also appreciates that he may intensify a defect by breeding to a dog that is good where his bitch is bad. For instance, if she is snipy in muzzle and of a family none too good in this point, and he finds a dog that is short and square in muzzle, and in fact very good in all points before the eyes, he does not jump at the conclusion that this is the dog for him to breed to. No. He goes carefully into his history, and if he learns that he comes from a bitch that was bad in muzzle and her family also had the same fault, while his sire was only fair in muzzle and of a family that were not noted for good muzzles, then he regards that dog as an "accident," and considers that were he to breed his bitch to him the existing defects would likely be intensified. Therefore, he seeks a dog that is known to transmit to or mark his puppies with good muzzles, and from parents or a family that were noted for the same excellent quality.