Therefore, he is wise who depends largely upon others for his sires. He will naturally have a leaning towards the most successful prize-winners, and rightly so, for in them he will generally find the nearest approach to the ideal, but he must not assume, as many seem to do, that these winners comprise all the best stock-getters. As a matter of fact no small proportion of them are practically valueless for breeding purposes. And of the various reasons for this fatal defect the most pronounced is that they are weakened by excessive use and the hardship they undergo in their preparation for shows and while on the bench; in consequence of which they are liable to produce offspring with constitutional taints that either destroy them early or lead to deformities and disease. Therefore, when selecting from prize-winners one must needs be very cautious, and bear in mind that high health and vigor are essentials of infinite importance.

The bitch is of course an important element, yet a good dog and a well-bred but rather indifferent bitch may breed the very best. In fact, there is reason to believe that the largest proportion of the winners have been from bitches that could not win.

Here intrudes another important fact, namely, that those who are breeding for size almost always select large bitches and large dogs, yet the majority of the big ones have been offspring of bitches of undersize that were bred to strong, heavy-boned dogs of substance.

As for in-breeding, as a rule it should be discouraged. But to produce certain results it is not only not harmful but justifiable; indeed, oftentimes it is the correct way to fix desirable qualities. For instance, both sire and dam may wisely be bred to their offspring, and even brother to sister, if by this means there is a promise that the higher qualities of each will be passed on and better development secured.

To justify in-breeding, however, both parties to the union must be of good quality and not share the same defects, and there must exist ample reason for expecting that the faults of one will be corrected or greatly lessened by the perfections of the other.

To breed as deeply as stated more than once in the same family without a break is doubtless at times justifiable; but still it is hazardous, and especially so unless the individuals are of rare good quality and remarkably healthy and vigorous.

The principal objections to in-breeding are, that it greatly weakens the nervous system, - producing excessive excitability, - intensifies constitutional defects, leads to decrease in size, creates a predisposition to disease and impairs the reproductive functions.

In a word, by the means of in-breeding desirable qualities may be made more pronounced, retained and perpetuated, but to be safe it must be very intelligently applied, with well-selected individuals only, and it must not go far in a family.

The conclusions to be drawn from the foregoing are, that the beginner should first visit shows and other places that furnish material, and there critically study the best specimens of his chosen variety of dogs. If closely observing he will in time become familiar with what are generally accepted as good qualities and be able to detect the faults. An ideal model will then be possible, and this developed and a knowledge of the common laws of inheritance acquired he will be fairly started and may hope to make breeding a success. But until he has reached this point he will do well to consult some acknowledged and disinterested authority in his line of dogs, and in the selection of sires act as he advises.

A very erroneous notion is prevalent regarding the influence of the previous sire on produce, and a hasty glance at the subject can properly be taken here.

Cases are on record in which bitches had litters by dogs of other breeds than their own and subsequent offsprings by dogs of their Own kind exhibited traces of the previous sires; and such instances, while extremely rare, have yet been too many and too well authenticated to be denied or explained away on the hypothesis of coincidence.

Various theories have been set forth as explanations of these exceptional phenomena, and as the writer has none to advance he will confine himself to mere review and passing criticism.

The first that found any considerable acceptance had its origin some twoscore years ago, but soon passed out of sight to be revived and brought into prominence by Sir Everett Millais of England, who gave it his indorsement.

This, in a nutshell, is, that the life-giving germ can penetrate the serous coat of the ovary, burrow into its parenchyma and seek out immature ova, not to, be ripened and discharged perhaps for years, and to produce the modifying influence described.

At the present time this theory is incapable of proof or disproof, but the careful student can but acknowledge that it is plausible, and after a consideration of the different . theories the author of this must seem to him to have at least approached nearer the border of the true solution of the vexed problem than any of the others who have attempted to reach it.

Another theory is, that the influence of the previous sire is due to maternal impression, consequently is purely mental.

This cannot be accepted. Puppies in utero are in every sense parasites and therefore beyond such influence. Again, if the phenomena were due to maternal impression all the members of a litter would likely be affected, whereas as a rule only one member exhibits influence of the previous sire.

Still another theory is, practically, that developing puppies in utero modify the tissues and cells of the mother, and such change is more or less permanent and affects subsequent offspring.

To believe this is to accept that a woman can be so modified in her physical texture by the impress of her husband, conveyed to her through the growing foetus, as to stamp the children by a future husband with the individuality of the former. Bearing in mind always that the child in utero is simply a parasite, and that while the developing tissues of the child may in a general way be influenced by the mother's condition and by poisons circulating in her blood, it is absurd to suppose that the parasite ovum can have a transforming and metamorphosing influence upon the already developed tissues of the mother. Moreover, were this theory sound the influence would be universally applicable instead of exceptional, as it clearly is. Again, were the influence transmitted through a change in the maternal make-up it would affect all members of a litter, and not, as is usually noted, only one or possibly two.