Sufficient materials do not exist for tracing the history of the separate breeds. About the commencement of the Christian era, Columella mentions a five-toed fighting breed, and some provincial breeds; but we know nothing about them. He also alludes to dwarf fowls; but these cannot have been the same with our Bantams, which, as Mr. Crawfurd has shown, were imported from Japan into Bantam in Java. A dwarf fowl, probably the true Bantam, is referred to in an old Japanese Encyclopædia, as I am informed by Mr. Birch. In the Chinese Encyclopædia published in 1596, but compiled from various sources, some of high antiquity, seven breeds are mentioned, including what we should now call Jumpers or Creepers, and likewise fowls with black feathers, bones, and flesh. In 1600 Aldrovandi describes seven or eight breeds of fowls, and this is the most ancient record from which the age of our European breeds can be inferred. The _Gallus turcicus_ certainly seems to be a pencilled Hamburgh; but Mr. Brent, a most capable judge, thinks that Aldrovandi “evidently figured what he happened to see, and not the best of the breed.” Mr. Brent, indeed, considers all Aldrovandi’s fowls as of impure breed; but it is a far more probable view that all our breeds have been much improved and modified since his time; for, as he went to the expense of so many figures, he probably would have secured characteristic specimens. The Silk fowl, however, probably then existed in its present state, as did almost certainly the fowl with frizzled or reversed feathers. Mr. Dixon[34] considers Aldrovandi’s Paduan fowl as “a variety of the Polish,” whereas Mr. Brent believes it to have been more nearly allied to the Malay. The anatomical peculiarities of the skull of the Polish breed were noticed by P. Borelli in 1656. I may add that in 1737 one Polish sub-breed, viz., the Golden-spangled, was known; but judging from Albin’s description, the comb was then larger, the crest of feathers much smaller, the breast more coarsely spotted, and the stomach and thighs much blacker: a Golden-spangled Polish fowl in this condition would now be of no value. _Differences in External and Internal Structure between the Breeds: Individual Variability._—Fowls have been exposed to diversified conditions of life, and as we have just seen there has been ample time for much variability and for the slow action of unconscious selection. As there are good grounds for believing that all the breeds are descended from Gallus bankiva, it will be worth while to describe in some detail the chief points of difference. Beginning with the eggs and chickens, I will pass on to their secondary sexual characters, and then to their differences in external structure and in the skeleton. I enter on the following details chiefly to show how variable almost every character has become under domestication.