ficial method is the one with which he is frequently forced to be content.

In actual practice the two systems are apt to be somewhat mixed. The mixture occurs because the pomologist seeks to make a natural classification; but, finding his knowledge imperfect and inadequate at certain points, is obliged to proceed without bringing positively all the characters of all the fruits into his scheme. As soon as he classifies anything on the basis of this partial knowledge, he is using a few characters, or a single character, in place of the full quota demanded by the perfectly natural method. Thus, we very commonly find the pomological worker piecing out a natural classification with artificial links.

The ideal system of classification combines both the natural and the artificial methods. If this ideal system were applied to any given material say, to our cultivated grapes-we should have two classifications given side by side. In the natural classification all the varieties would be arranged according to their broad, natural relationships. In the arbitrary classification the same series of varieties would be arranged more diagramatically

according to characters arbitrarily chosen. Then any one wishing to identify an unknown variety could follow out the arbitrary key or classification, while any one wishing to study the varieties broadly would have recourse to the natural classification.

It should be remarked, for the benefit of those who are accustomed to the use of the current manuals of botany, such as Gray's Manual, that these books employ, more or less successfully, the double method outlined. The keys given in these books for help in tracing out (" analyzing," as they say in school) plants whose names are not known are purely artificial. The arrangement of plants into species, genera, and families, according to their broad resemblances, is natural—at least, it is professedly so; and though the books frequently fall far short of the whole truth in these qualifications, yet the arrangements are on a natural basis.

All these matters will appear more clearly when we take up some of the actual examples of classification, as we shall now proceed to do. The author feels that some explanation is fairly due the student for the introduction of the following very imperfect, and often

contradictory, systems of classification. It would be much better if this treatise on systematic pomology could give an ideal classification of our common fruits. Such a classification, however, has not yet been made, and it seems impossible that it should be madeat least, for many years to come.

In default of a perfect classification, therefore, the reader will be instructed by observing what the best pomologists have already done in this field. Each one of us will then be at liberty to choose for his own use that system of classification which seems to him most reasonable or useful. In many

cases, doubtless, two or more systems of classification can be combined, or useful features can be taken from each. Out of such study, readaptation, and invention the future may see a better system of pomological

classification arise.

XIV

CLASSIFICATION OF FRUITS IN GENERAL

IT is difficult to define a fruit in exact terms. The botanist has a definition, but it is not just the same as the horticulturist's definition. In general, we may sum up the horticultural notion by saying that a fruit is an edible, more or less fleshy portion of a plant, in its development intimately connected with the seed. Sometimes it is the seed itself, as in the walnut; sometimes it is the swollen ovary, as in the plum; sometimes it is the fleshy calyx adhering to the ovary, as in the apple.

Such

The names of fruits in common language are really generic classificatory terms. names as peach, nectarine, apricot, almond, etc., designate rather considerable classes of fruits. While such names are usually fairly clear in common usage, it is sometimes difficult, in critical cases, to say just what is a plum and just what is a cherry. The shrub commonly called the western (Prunus pumila besseyi) has

sand cherry often been

classed with the plums. It may be called by either name with equal propriety.

Recently the plum has been successfully hybridized with the apricot and with the cherry. The resulting fruit is half plum and half apricot, and can not be conveniently classed with either. Mr. Burbank has called it a plumcot. The cross between the plum and the cherry has not been named.

Whether such pears or apples

The writer has several times seen fruits which were thought to be hybrids between the apple and the pear. They partook of the characters of both fruits. specimens could be called would be a knotty question. thing that can be done in such cases is to rely on the arbitrary definitions of our common language.

About the best

When we study all these common classes of fruit at large, we find that there are some striking natural relationships among them. The apple, the pear, and the quince, for instance, are much alike; so are the plum, the peach, and the apricot; and so are the orange, the lemon, the pomelo, and the kumquat. The three groups here exemplified are very generally known as the pome fruits,