In 1887 Gilbert Onderdonk,[185] a special agent of the United States Department of Agriculture, published a natural classification of peaches.[186] He put varieties of peaches into five groups which he called races and to which he gave the names: Persian, Northern Chinese, Spanish, Southern Chinese and Peento. He bounded peach-culture in America on the north by the Great Lakes and on the south by the Gulf and divided this great region into five zones to each of which he assigned one of his races. Onderdonk studied peaches in Texas and found there remarkable distinguishing characters; as, in adaptations to southern climates, in length of the rest-period, in differences in leafing, blooming and fruiting-time, and in the organs of the plants. Professor R. H. Price, working with a large number of varieties at the Texas Agricultural College, verified and greatly extended Onderdonk's observations.[187] Eventually, Price became the pontifical authority in this country on the classification of peaches and in numerous articles and addresses set forth the Onderdonk grouping of varieties so convincingly that it was adopted by practically all American pomologists and at present is in use, to some degree at least, in nearly all of our horticultural literature. It becomes necessary, therefore, to scrutinize closely this natural classification of Onderdonk and Price. The end to be attained in a classification of peaches, as in classifying natural objects of any kind, is to provide an epitome of the knowledge of the fruits classified. Incidentally, a classification helps in the identification of varieties of peaches. Does the Onderdonk classification serve these purposes? We have not found that it does. In most arduous attempts to arrange the sorts of peaches growing on the Station grounds according to the Onderdonk plan, we have wholly failed. Even the varieties named as types do not fit, as they grow in the north, in the places provided for them by these southern classifiers. Indeed, we have wasted so much time and patience in attempting to group varieties according to Onderdonk and Price, and with so little success, that the Onderdonk classification seems to us to be cursed with the confusion of Babel. Since pomologists so generally accept this classification, these words demand that it be shown wherein this attempt at a natural arrangement of varieties fails.