The stripes are variable in colour, but are always darker than the rest of the body. They do not by any means always {60} coexist on the different parts of the body: the legs may be striped without any shoulder-stripe, or the converse case, which is rarer, may occur; but I have never heard of either shoulder or leg-stripes without the spinal stripe. The latter is by far the commonest of all the stripes, as might have been expected, as it characterises the other seven or eight species of the genus. It is remarkable that so trifling a character as the shoulder-stripe being double or triple should occur in such different breeds as Welch and Devonshire ponies, the Shan pony, heavy cart-horses, light South American horses, and the lanky Kattywar breed. Colonel Hamilton Smith believes that one of his five supposed primitive stocks was dun-coloured and striped; and that the stripes in all the other breeds result from ancient crosses with this one primitive dun; but it is extremely improbable that different breeds living in such distant quarters of the world should all have been crossed with any one aboriginally distinct stock. Nor have we any reason to believe that the effects of a cross at a very remote period could be propagated for so many generations as is implied on this view. With respect to the primitive colour of the horse having been dun, Colonel Hamilton Smith[136] has collected a large body of evidence showing that this tint was common in the East as far back as the time of Alexander, and that the wild horses of Western Asia and Eastern Europe now are, or recently were, of various shades of dun. It seems that not very long ago a wild breed of dun-coloured horses with a spinal stripe was preserved in the royal parks in Prussia. I hear from Hungary that the inhabitants of that country look at the duns with a spinal stripe as the aboriginal stock, and so it is in Norway. Dun-coloured ponies are not rare in the mountainous parts of Devonshire, Wales, and Scotland, where the aboriginal breed would have had the best chance of being preserved. In South America in the time of Azara, when the horse had been feral for about 250 years, 90 out of 100 horses were "bai-châtains," and the remaining ten were "zains," and not more than one in 2000 {61} black. Zain is generally translated as dark without any white; but as Azara speaks of mules being "zain-clair," I suspect that zain must have meant dun-coloured. In some parts of the world feral horses show a strong tendency to become roans.[137]