XIII PRINCIPLES OF CLASSIFICATION THE natural end of systematic pomology is classification. "Science is classified knowledge," in the terms of the old school-book definition. Until we classify our pears, plums, raspberries, and other fruits, our knowledge of them does not really become scientific. Description and nomenclature are often valuable in themselves, but they do not reach their legitimate application until they are made the means to a classification. The present need of more and better classification in pomology deserves to be very forcibly presented. Horticulturists will agree, on general principles, that our fruits deserve as careful study and as accurate classification as the different kinds of microbes floating in the air, or the various sorts of moss clinging to the rocks of Popocatapetl. Pomology is just as promising and as proper a field for science as bacteriology, mycology, or conchology. Then when one remembers that there are probably be tween 1,500 and 2,000 different named varieties of apples growing in North America, some 500 to 600 varieties of plums, 200 to 300 of and of other fruits in like proporpears, tion, with more coming every year, then one must see that, unless some method of topical study can be secured, no man can ever get even a superficial knowledge of our pomological wealth. If those varieties which are alike can be grouped together, then one can understand several of them at once by familiarizing himself with their generic characters. And thus from every side, and much more urgently than can be here written down, do we need more and better classification of our varieties of fruit. Methods of Classification There are two radically different methods of classification used in all fields of natural science. These are commonly designated the natural and the artificial methods. It would be more accurate and more suggestive to call the latter the arbitrary method rather than the artificial method. The essential difference between these two methods is that the arbitrary method arranges objects (or varieties in this case) into groups according to some one character arbitrarily selected, while the natural method attempts to take into consideration all the characters of the varieties in question. This matter may be made plain by an example. In nurserymen's catalogs apples are usually classified according to one character— namely, season of ripening; that is, they are given a purely arbitrary classification. There may be two varieties almost exactly alike as regards growth of tree, color, texture, and flavor of fruit, and all other qualities; yet if one happens to ripen in August and the other in December, the two will go to the opposite ends of the nurseryman's list. We would have similarly an arbitrary classification if we selected any other one character for our basis of assortment. We might decide to classify according to form, which is one of the best of all arbitrary ways of classifying apples. We would put the round apples into one group, the oblate apples into another, the oblong varieties into a third, and the conic varieties into a fourth. Still we would be sure to find placed together certain apples which, though of the same form, were very unlike in most other characteristics; and conversely we would discover that many varieties, though strikingly alike in the majority of their characters, had been separated into distinct groups by our arbitrary method of classification. To illustrate the natural method of classification among apples we may refer to the fairly well-recognized Fameuse group. This includes Fameuse (Snow), McIntosh, Shiawassee, Sweet Fameuse, Louise, Green Fameuse, and others all having certain points of resemblance. They are usually red (but not always); they have tender skin, soft, tender, very white flesh, with pinkish markings running through it, a peculiar flavor and aroma, and a medium early season of ripening. The Fameuse group, therefore, is founded on natural resemblances. In determining these resemblances the apples are considered in all their characters, not in some single one. Persons who are not familiar with apples of the Fameuse group may understand this point better by calling up the Ben Davis group. This consists of Ben Davis, Gano, Black Ben Davis, Beach (Apple of Commerce), and several others—all much like Ben Davis in all their principal qualities. It will be seen at once that the arbitrary method of classification is the easier to use. It is easier to study varieties for agreements or disagreements in one character than in many characters. The arbitrary classification is also easier to make, as well as easier to use. Any natural classification is extremely difficult with any kind of fruits, but especially so with apples, pears, and strawberries. In fact, with these fruits it is practically impossible to make a natural classification covering more than a few specially well-defined groups. The natural method of classification, however, is the better for all reasons, except that of convenience. It is more scientific-if that adjective permits of comparison-than the arbitrary method, because it shows more of the facts of relationship. Fruits which resemble each other in several particulars are certain to be more closely related, in the majority of cases, than those which resemble only in one particular. The natural method is, therefore, the one toward which the ambitious pomologist always strives, but the arti |