78 TRANSACTIONS OF THE ILLINOIS 



rials from roots of another variety, is no reason why the eion should 

 partake of the characteristics of the root. Neither is it possible, 

 under the general law of inheritance, for such a root to take the 

 characteristics of the cion because the leaves of the latter elaborate 

 and send down food for the sustenance of the root. If there is any 

 blending whatever in the peculiar characteristics of the cion and 

 the stock it must be by a union of the semi-solid protoplasms of both. 

 This is just what takes place in sexual propagation. In the case of 

 flowers a part of the protoplasmic contents of the pollen grain 

 actually reaches and coalesces with that of the germ cell of the 

 embryo, and the latter partakes of the nature of both elements. 

 From the very beginning this embryo is nourished by the pistillate 

 or mother plant, yet we see no predominance of this latter over the 

 pollen-bearing or male plants. The controlling impress is upon the 

 original protoplasmic parts, and this persists through the life and 

 growth of the individual, with only such variations in size and vigor 

 as difference in nutrition may make. It is well known that cross- 

 fertilization can take place only within very narrow limits of struc- 

 tural variation. Even with very closely allied plants, as the pear 

 and the apple, it cannot be foretold whether or not hybrids can be 

 produced, though experience has made us confident that these can 

 only arise between closely related species. 



If now, in the process of grafting similar protoplasmic union 

 does at all take place, we shall look for results closely approximating 

 those of normal sexual combination. The question of graft hybrids 

 really lies before that of the influence of the cion on the stock and 

 vice versa. Let us carefully note the difference between simple cellular 

 union, by which a cion may live upon a given stock, and true cross- 

 ing. In the ordinary methods of grafting the united surfaces are 

 more or less transverse cuts, but there is no reason why the same 

 cellular union should not take place with longitudinal cuts. When 

 it is asserted that the terminal buds of red and blue hyacinths may 

 be split, and the half of one combined with a half of the other, and 

 thus combined grow into a stem with the characteristics of both, 

 this is only grafting with the usual union of cell tissue, and not 

 properly the production of a graft hybrid. Each half of the stem 

 keeps its proper peculiarities, and the only difference from the results 

 of ordinary grafting, is that here the stock and cion grow side by 

 side, instead of one above the other. In a true hybrid, from graft- 

 ing as well as from pollen fertilization, the characteristics of the two 

 parental forms are blended through and through in the offspring. 

 Remembering the process of sexual union in the flower, no one can 

 assume that the same thing may not take place as a result of graft- 

 ing. Most assuredly in the fitting of the cion to the stock the naked 

 protoplasm of the two meets and, for all we know, may physiologi- 

 cally combine. If so, there is only needed a bud from the combined 

 portions to give us a true graft hybrid. There is good evidence that 



