nating are but recombinations of the characters in the parent—the combination is new but not the characters. Thus one parent of a hybrid grape may contribute color, size, flavor and practically all of the characters of the fruit and the other parent vigor, hardiness, resistance to disease and in general the characters of the vine. Or, of course, these and the other items in the make-up of the grape may be intermingled in any mathematically possible way. New characters probably appear as variations, and of these plant-breeders now recognize two kinds.
Nothing is more certan than that all offspring differ from their parents in many details—individual variation. Plant-breeders have long believed that by selecting desirable variations we have an efficient means of improving plants just as evolutionists have held and many continue to hold that evolution goes forward by means of natural selection from these variations. But there is a new school, headed by the Dutch botanist, De Vries, who believe that these variations do not produce anything new, but that they always oscillate around an average, and if removed from this for a time, they show a tendency to return to it. Whether the orthodox Darwinians or the De Vriesians are right does not matter here. The point is that the fluctuating variations of individuals, upon which Darwin chiefly founded his principle of natural selection, cut but a small figure in the breeding of grapes. It is not certain that such, variations are heritable, nor whether they are capable of cumulative increase generation after generation, and, besides, as we have seen, selection must be consistent and persistent for too long a while to make it effective with grapes.
Evolution and plant-breeding have taken a fresh start through the recent amplification by De Vries of the theory that marked changes take place in plants through mutations, or characters which arise in a plant at once, with a single leap, and are stable from the time they arise. If this theory hold for grapes, it may be that there is a possibility of absolutely new characters arising in this fruit. It is well known that bud-sports, which in most cases must be called mutations, now and then arise in grapes. But these mutations have not as yet played an important part in producing new varieties. Not more than two or three of the fifteen hundred sorts now under cultivation are suspected of having arisen in this way. Until the causes of these mutations are known and they can be produced and controlled, but little can be hoped for in the amelioration of grapes through mutations.
Hybridization, then, has been and continues to be the chief means of domesticating grapes. "Fluctuations" and "mutations," produced other than by hybridizing, are too vague as yet for the grape-breeder to lay hands on. Even should the theory of De Vries be true, that nothing new—in the strict sense of the word—comes except through mutations, with more than a score of species of grapes, each with manifold distinct characters, all capable of fluctuating variations, there are many
