210 A TEXTBOOK OF BOTANY [Ch. IV, 13 



grafting may be defined from this point of view as a process 

 of fitting a set of ready-made roots upon kinds of plants 

 unable to make any of their own. 



Second, grafting can be used to produce certain desirable 

 changes in minor qualities of the cion, though no essential 

 features can thus be altered. An earlier or later time of 

 blossoming or fruiting of a tree, a better adjustment to a 

 particular soil or climate, advantageous dwarfing or enlarg- 

 ing, resistance to root parasites, even in some small degree 

 an improvement in color or size may be wrought in the cion 

 by grafting on a suitable stock. All such features, however, 

 seem to depend upon the sap, which of course is supplied by 

 the roots of the stock. The more essential characters are 

 seated in the protoplasm, and remain unaltered by grafting, 

 since the protoplasm, unlike the sap, does not pass from 

 stock to cion, but remains separate in the two. 



Third, curious effects in plant form are obtainable by 

 grafting, as when a dozen or more varieties of Cherries are 

 made to grow on one tree, or bizarre constructions are pro- 

 duced by the grafting upon one stock of many forms of Cacti, 

 which happen to graft extraordinarily well. 



The older books upon horticulture frequently mention 

 graft-hybrids, of which the most famous is Cytisus Adami, 

 produced by grafting between yellow-flowered and purple- 

 flowered shrubs, and itself preserved by grafting. It shows 

 diverse comminglings of yellow and purple in the flowers, 

 but not an intermediate color. In a true hybrid, produced 

 by the crossing through fertilization of two parents of dif- 

 ferent races or species, the color is that of one parent or the 

 other, or else has an intermediate shade, but is never a 

 mosaic of the two colors, as in this plant. However, modern 

 research has shown that Cytisus Adami is no hybrid at all, 

 but a mixture of the tissues of the two parents, such a com- 

 bination being now called a chimera. It has been found 

 possible to produce these chimaeras artificially by so manip- 

 ulating the grafting that a part of a bud of the cion unites 



