BOOK XVII. XXIV. 102-105 



})ith would not take more. But subsequently a 

 more elaborate method is for as many as six grafts to 

 be added to rcinforce their liability to die and their 

 number," a cleft being carefullv made through the 

 middle of the trunk and being kept open by means 

 of a thin wedge until the graft, the end of which 

 has been pared into a point, goes right down into 

 the crack. 



In this process a great many precautions have to be 

 observed. First of all we must notice what kind of tree 

 will stand grafting of this nature, and what tree it will 

 take a graft from. Also the sap is variously distri- 

 buted, and does not He under the bark in the same 

 parts with all trees : in vines and figs the middle is 

 drier, and generation starts from the top, shoots for 

 grafting being consequently taken from the top of the 

 tree, whereas in olives the sap is round the middle and 

 grafts are also taken from there, the tops being parched 

 up. Grafts and trunk grow together most easily 

 when they have the same kind of bark and when they 

 flower at the same time, so that they have the affinity 

 of the same season and a partnership of juices ; where- 

 as it is a slow business when there is incompatibiHty 

 between dry tissues and damp ones, and between hard 

 and soft barks. The other points to be observed are 

 not to make the clcft at a knot, as the inhospitable 

 hardness repudiates a new-conier; to make it at the 

 shiniest place ; not to make it much more than three 

 inches long, nor on a slant, nor so as to be transparent. 

 Virgil'' says that grafts must not be taken from the 

 top, and it is certain that the sHps should be obtained 

 from the shoulders of the tree that k)ok north-cast, 

 and from trees that are good bearers and from a young 

 shoot, unless the tree on which they are to be grafted 



73 



