RACES are such strongly marked varieties, capable of coming true to seed. The different sorts of Wheat, Maize, Peas, Radishes, etc., are familiar examples. By selecting those individuals of a species which have developed or inherited any desirable peculiarity, keeping them from mingling with their less promising brethren, and selecting again the most promising plants raised from their seeds, the cultivator may in a few generations render almost any variety transmissible by seed, so long as it is cared for and kept apart. In fact, this is the way the cultivated domesticated races, so useful to man, have been fixed and preserved. Races, in fact, can hardly, if at all, be said to exist independently of man. But man does not really produce them. Such peculiarities--often surprising enough--now and then originate, we know not how (the plant _sports_, as the gardeners say); they are only preserved, propagated, and generally further developed, by the cultivator's skilful care. If left alone, they are likely to dwindle and perish, or else revert to the original form of the species. Vegetable races are commonly annuals, which can be kept up only by seed, or herbs of which a succession of generations can be had every year or two, and so the education by selection be completed without great lapse of time. But all fruit-trees could probably be fixed into races in an equal number of generations. BUD-VARIETIES are those which spring from buds instead of seed. They are uncommon to any marked extent. They are sometimes called _Sports_, but this name is equally applied to variations among seedlings. CROSS-BREEDS, strictly so-called, are the variations which come from cross-fertilizing one variety of a species with another. HYBRIDS are the varieties, if they may be so called,--which come from the crossing of species (331). Only nearly related species can be hybridized; and the resulting progeny is usually self-sterile, but not always. Hybrid plants, however, may often be fertilized and made prolific by the pollen of one or the other parent. This produces another kind of cross-breeds. 525. Species are the units in classification. Varieties, although of utmost importance in cultivation and of considerable consequence in the flora of any country, are of less botanical significance. For they are apt to be indefinite and to shade off one form into another. But species, the botanist _expects_ to be distinct. Indeed, the practical difference to the botanist between species and varieties is the definite limitation of the one and the indefiniteness of the other. The botanist's determination is partly a matter of observation, partly of judgment.