Going still further, we may compare the effects of cross-breeding or of hybridisation, where the progeny show that changes have resulted from the mutual interactions and reactions of the commingled protoplasm, with Daniel's results, in which he obtains proof of such interactions of the commingled protoplasmic cell-contents of grafts in the seedling progeny; although there is no probability - we may even say possibility - in this latter case that the effects are due to nuclear fusions, but only that the germ-plasm of the seed-bearing plant has been affected by the changes in the cell-protoplasm which nourishes it when the reproductive cells are forming.
In the case of graft-hybrids the matter appears to be somewhat different, and we may well suppose, with Strasburger, that the commingling of characters observed in flowers, fruits, foliage, etc., on shoots borne after grafting are due to the occurrence of nuclear fusions during the union of the grafted tissues; though it is by no means impossible that what has really happened is profound alterations in the nuclear substance (germ-plasm) owing to its being nourished by cell-protoplasm (somato-plasm) which has been itself affected by the interchanges of substance between scion and stock, and therefore itself furnishes a different nutrient medium from the unaltered cytoplasm of either.
But even here we can find parallels among the ordinary phenomena of plant reproduction. Maize plants with white endosperm containing starch, if crossed by pollen from other plants with purple endosperm containing sugar, bear seeds with purple endosperm containing sugar, and such Xenia may be compared to graft-hybrids in many respects.
I know of no case among fungus infections which could be compared directly with these examples, and it is not at all likely that we shall meet with any instance of a fungus-hypha handing over nuclear substance to an egg-cell, and so affecting the latter that an embryo results.
But the case is not hypothetically impossible, although the distant relationships of the two groups of organisms render it extremely improbable among the higher plants. It is by no means so improbable, however, that further research may show cases where the egg-cell of a lower cryptogam - e.g. another fungus - may be affected either directly, or indirectly, by the protoplasm of a parasitic or symbiotic hypha, as suggested by the extraordinary phenomena of symbiosis.
Some of the variations in grafted plants are found to predispose the plant to disease, or the reverse, and cases may be cited where the resulting shoots, foliage, or fruits, or seedlings more readily fall a prey to, or resist, parasitic fungi and insects than the ungrafted plants. Daniel gives instances of such - e.g. among other examples, Peas grafted on Beans yield seeds which suffer more from Erysipheae than the normal seedlings. But the best known cases are those of Vines in their relations to Phylloxera, already referred to (P. 155).
Several instances are also known where grafted plants show more or less resistance to such factors of the environment as low temperatures; grafted or budded Roses often suffer much from Erysipheae, and so forth. Much research is still needed to determine how far these matters depend on real alterations in the nature of the graft, or are only true for the localities in which the experiments have been made, a point which has, I think, been overlooked by all observers.
Grafted plants are apparently very much exposed to injury by slugs, insects, and the invasions of parasites during the healing of the callus and the fusion process. Here again it must not be overlooked that the callus is, so to speak, a tit-bit of luscious, thin-walled, succulent tissue; and, like all wounds, the graft affords entrance to parasites such as Nectria and Ascomycetes of various kinds, under circumstances very favourable to their invasion.
Natural Plant Grafts. - It is by no means an uncommon event to find the branches of Beeches, Limes, and other trees which have been accidentally brought into contact during growth, joined where they cross. As they press one against the other, they become naturally grafted, by that form of the process known as inarching: except that in artificial inarching the operator cuts off the cortical tissues of the two branches and brings their cambial surfaces together, whereas in nature the cambiums only come into contact after the destruction by pressure, or slight abrasion, of the entrapped intervening tissues. The fusion occurs, in fact, exactly as in the burying - in of a nail or wire, referred to on p. 211.
Natural grafts are very common among the roots of trees, and possibly explain some queer cases of the apparent revivification of stumps of trees not usually given to forming abundant stool shoots. It is regarded as probable in some old forests that the majority of the roots of trees of the same species are linked up together by such natural grafts, a probability not diminished by the fact that such roots cross at many points, and are easily grafted.
The student should read Bailey, The Nursery Book, 1896, for details regarding the practice of grafting, and facts in abundance can be obtained from the pages of the Gardeners' Chronicle.
Concerning graft-hybrids and the variations of grafted plants see Jouin, Can Hybrids be obtained by Grafting? and especially Daniel, "La Variation dans la Greffe," in Ann. des Sc. Naturelles, S. VIII., Vol. 8, 1898, p. 1, and the literature there collected. The whole subject is largely controversial, and much work remains to be done.