74 SEEDLING PEOPAGATION OF VARIETIES. 



in attaining perfection, but impracticable on a large 

 scale. 



METHODS OF GRAFTESTG. 



Scions for grafting should be of one or two years' 

 growth, that have not yet produced fruit-buds. The 

 shoots selected should be firm-wooded and stocky, 

 with buds close together, as a strong, healthy growth 

 is characterized by these marks. 



Grafts taken from the upright shoots near the top 

 of the tree are apt to make a vigorous and upright 

 growth, but are more tardy in bearing. Taken from 

 the lower part of the tree, they produce a more widely- 

 spread form, and fruit earlier. 



The trees from which the grafts or buds are taken 

 should be healtliy, and have produced a vigorous 

 growth during the previous season, but such as have 

 at any time exhibited symptoms of frozen sap-blight 

 should especially be avoided. 



Yarieties which succeed but indiiferently on quince 

 stocks, ought not to be propagated by scions from 

 trees grown on quince. Indeed, it is a mooted ques- 

 tion whether grafts should be taken at all from such 

 a source ; but I see no reason for going to this extreme. 



The j)art of the graft used with the m.ost success, is 

 that at the junction of the spring and midsummer 

 growth, which is marked by a somewhat fainter annu- 

 lar swelling than that at the commencement of the 

 spring growth. 



The theory of grafting is, that the newer tissues of 

 woody growth unite, when brought into contact, if 

 their sap-vessels are not indurated by age. The ter- 

 mini of the cellular tubes are caj^able of exuding 



