840 



of the swollen walls of the outermost cells, or caused by the collapse of all 

 the cells of the wound edges. Usually on the boundary a cork layer has 

 been formed by the suberization of the walls of the peripheral parenchyma 

 cells or, besides this, by the appearance of actual cork cells. 



In genera which finally unite, as, for example, Iresine on Alternanthera, 

 it is found that for whole stretches of the grafted surfaces, the connecting 

 tissues grow side by side, cut off from each other by a cork layer. 



Similar cases may be proved in root grafts (Bignonia) and it could be 

 observed in cleft grafts of Paeonia arhorea on fleshy roots of Paeonia offici- 

 nalis that the root, as stock, had served only as a receptacle for the scion. 

 This latter had formed roots without any union with the stock. 



Root grafting is in general a very good method. Even for our fruit 

 trees it had been used by Sickler at the end of the seventeenth ( ?) century 

 and later Seigerschmidt in Mako recommended it very highly \ Root 

 l)ieces, varying in thickness from the size of a quill to that of one's thumb, 

 seemed suitable, if provided with fine roots. They were cut in pieces eight 

 to twelve cm. long, were grafted by copulation or cleft grafted and the place 

 of union covered with earth until only two or three eyes extended above the 

 soil. Trunks of old seed- or stone-fruits give an abundant material for 

 stock, when they have to be removed. Of course, the roots must be very 

 healthy. The method of grafting roses on pieces of roots in January or 

 February has been adopted already. For Clematis and other woody plants, 

 this method of grafting is becoming more and more of a favorite. 



It may be presumed from the very beginning that under certain cir- 

 cumstances which condition a scanty coalescence, the life period of a graft 

 will be very short. The question, whether the process of grafting in itself 

 limits the life period, as Thouin and Goppert have stated, must be laid 

 aside. . It cannot be denied that grafted fruit trees, on an average, are 

 shorter lived than those grown- on their own roots. It may even be granted 

 that a dying of the trees, as Goppert has observed, is initiated in the line of 

 demarcation by a gradual rotting of the place of union but it is not credible 

 that this process of rotting may be the cause of actual death, or even of 

 disease, in grafted trees. It is found, on the contrary, that even badly united 

 copulants, which at first may have been simply stuck together on one side, 

 can in the end give perfectly healthy permanent trunks. The old places of 

 union have the firmest wood. A storm may twist the trees off more easily 

 at any other place than at that of the union. Goppert's observations may 

 possibly hold as the rule only in old trunks which have been regrafted later. 

 I would explain the comparatively earlier death of grafted trunks by the 

 fact that not only better but also more tender cultural varieties are used for 

 grafting. These, aside from the disturbances which they undergo in the 

 cutting, are in themselves more susceptible to disturbances in growth and 

 to unfavorable weather than the specimens grown from the seed, which 

 approach more or less the hardier nature of the stock. 

 1 Weiner Obst- und Gartcnzcitung: 1876, p. 587. 



