In breeding for the shortest faces we can get, instead of accepting anything and everything which has no bridge to its nose and indiscriminately making champions of all the ragtag and bobtail of the Bulldog cross, let us bear in mind the laws which should govern the short-nosed Toy type and breed Toys with the pretty expressions that come only with harmony of line and the observance of mathematical canons of proportion, I was the first to offer prizes for expression, as also for action, and there has since been a lot of talk about both, but some of the fanciers who have taken up the catchword most in the press do not understand expression as I understand it. They overlook things that I hold in abhorrence - oblique eyes, hollow cheeks, and

1 It will be seen that throughout this work I shall refer to the sexes as "dog" and "bitch" respectively. This would hardly seem to call for comment but for the habit which "lady breeders" have of referring to their dogs as "the little lady" or the "little gentleman," " my little boy " or "my little girl," to the great bewilderment of the uninitiated, and to crown all, the new-fangled word "matron" has been introduced among the would-be genteel. I did once hear of a society lady who referred to a foal as a "horse child," but I have not heard that the expression has come into general use among horse breeders, nor do they refer to their mares as "matrons!" Yet; as I have said before, we talk of dog-children.

I greatly object to the spurious gentility which makes impropriety where none exists, so I shall use plain unvarnished English words. Honi soit qui mal y pense! muzzles that are far too deep from nose to under jaw.

So that there may be no misunderstanding as to my views on the subject, please refer to the coloured plate. This is my idea of the ideal we have to keep in mind for a short nose. Look at the cushioned muzzle, the lovely low-set eyes; the round skull, and the sweet expression. Compare it with the other heads in this book point by point. Make up your mind to see where the differences lie and why it is pretty, while equally short-faced dogs are often so ugly. It is not the result of chance, and if you study it enough you will never again be able to tolerate the ugliness which fills our Show benches.

I propose to show the best way to breed the ideal short-nosed head, as I understand it, and, with the marvellous facility shown by our breeders in producing what they are trying for, as exampled in some of the amazing deformities of our day, I am sure that when they try for what is the right type they will very soon get it

Remember that the curves of the head and face must harmonise one with another and must not be violently intersected by meaningless angles and irregularities. Lips must not be jagged in outline, but the curve must be perfectly soft and even, as though drawn with one sweep of the brush. The skull must also be in a firm, circular curve, unbroken by peaks, knobs, or depressions of any sort. The cushions of the muzzle must be likewise semi-circular in outline. The eyes must be large, wide, and full, but not goggled, like those of a man with Graves's disease. The under eyelids have the same marked curve. The neck is arched, and the circular system is carried on by the markings on the head and the spot in the broken colours and in each variety by all the lines of the face.

The Blenheim should never compete against the Black-and-tan. It is quite absurd. A Black-and-tan fancier would give the prize to a bad Black-and-tan against a perfect Blenheim ninety-nine times out of a hundred, as also would a variety judge.

The breeds are emphatically not the same, and, just as a Black-and-tan is never got from two Blenheims, so, as a general law, a Blenheim will never be bred from two Black-and-tans. Neither will a Tricolour be bred from two Black-and-tans or two Blenheims. Rubies, however, will appear - so I am told - from two Black-and-tans, though I have never bred one this way myself, but I do not think two Rubies ever bred a Black-and-tan - certainly never within my observation. Mr. Milnes has given me the only instance on record of two Black-and-tans alleged to have bred a Blenheim, and vouches for its authenticity. In the case claimed there was a cross of Blenheim in the immediate ancestry of both sire and dam.1 The crossing of the breeds produces in the first generation a mixed type and colour. It will be noticed that from this mixed type are bred the

1Bcfore the case could be accepted as scientifically proved it would be necessary to eliminate all possibility of a mistake and there are too many unknown factors to make this a conclusive instance. I have had only two cases submitted to me of two Tricolours alleged to have produced a Blenheim. Of these one of the pedigrees is based on a dog that is notorious as a non-stockgetter and the other is equally unreliable as the identity of the sire is more than doubtful.

I have, however, come across an authentic case of two blue-and-tan puppies bred from black-and-tan parents. It is curious to note that a mis-marked Ruby if whole-colour bred will get all mismarked puppies to a Blenheim just as if she were whole coloured herself.

Ruby on the Blenheim side and Tricolour on the Black-and-tan side, and that both these varieties breed true to the new colour when bred each one to itself, but when bred together revert to the mixed type and colour. Each of these varieties bears a stamp of its own. The Tricolour is nearest the Blenheim in conformation, and the Ruby approximates the Black-and-tan, though neither has quite the same type as either grandparent. This is very curious indeed, as it would seem that the colour and the general conformation go by inversion, each grandparent supplying one of the main characters, the Tricolour getting the black from the Black-and-tan and the lighter type from the Blenheim, and the Ruby getting the red from the Blenheim and the heavier type from the Black-and-tan. Both, however, are inclined to reproduce the noseless head, and the Blenheims of the same generation also reproduce it, and take the heavier type.