In the first place the basis of Onderdonk's classification, as the names suggest, is regional variation. Each race stands for a region, the Peento included--for the name is very obviously Chinese. Incompleteness, then, is the first fault of this system for there are other regions in which races of peaches just as distinct as those named have developed: as, for examples, the Bokhara represents a hardy "Russian race;" Yellow Transvaal belongs to the very peculiar "South African race;" in the rich alluvial lands of Egypt, the "Egyptian race" has developed; still another regional race is found in the evergreen peach of the West Indies. We have no doubt that distinct races of peaches may have originated or will arise in the Canary Islands, Hawaii, New Zealand, Argentina, Chili and Mexico, to mention only countries spoken of in the foregoing pages. The Onderdonk classification can, of course, be extended to take in these new races, most of which are now represented in America, but eventually such a classification would become too cumbersome for use. It must not be overlooked that the Onderdonk classification should be doubled to apply to the nectarine, the other division of _Prunus persica_, which the present classification wholly ignores. If the variations are stable, and all regions represented, the likenesses and differences brought about by regional environment may well be used by classifiers. But in the Onderdonk classification unstable variations due to climate are too largely used; as, differences in the succession of life-events, in the rest-period, in the capacity to endure heat and drought, and in minor modifications of organs, as color of foliage and shape of fruit. All of these are variations that fluctuate with even slight changes in the climate. We have said that this classification, though constantly referred to by northern fruit-growers, is not satisfactory in New York. Professor Price, too, found as he went northward that his classificatory scheme was less dependable. He says:[188] "Some of the distinctions made in this classification cannot be noticed with decisive clearness a few hundred miles farther north." A further objection to this regional classification of Onderdonk is that, in the numerous distinct peach-regions of America, new regional variations are arising which make it impossible to classify in accordance with characters that appeared before the peach came to America.