THE PROPAGATION OF APPLE TREES. 77 



teresting to review after eight years' work on the problem. The first 

 question was, "Have you ever observed root-gi-afted apple trees rooting 

 from the scion?" Fifty replies say yes, and 6 reply no. Especially in 

 the Middle West nurserymen regard it as a common or usual thing, while 

 in the East, South and on the Pacific coast it seems rather less well known. 

 It may be that rooting is more frequent in the rich, loamy soil of the 

 Middle West, or it may be that it is because the practice of root grafting 

 prevails there more than in the eastern and other nursery regions. 



The second question asked, "In what varieties, and in about what 

 proportion of the trees," rooting from the scion had been observed to 

 occur. The general trend of the replies was that all varieties might do so, 

 Winesap being the only sort mentioned as not rooting. Generally the 

 varieties mentioned were those most extensively grown. Ideas as to pro- 

 portion of trees rooting were diverse, some saying a small percentage and 

 others nearly all. 



A question as to the most favorable conditions for rooting brought in 

 nearly every case, when a positive reply was made, the suggestion of the 

 long-scion, short -root graft; deep planting was often suggested; abun- 

 dant fertility and plenty of moisture were often mentioned; where soil 

 preference was expressed it was for a sandy or loamy soil. 



Methods used. 



The first lots of grafts for the purpose of securing trees on known roots 

 were made in 1912, and others were made during subsequent years, in- 

 cluding 1917. The method has been to make an ordinary piece root, 

 whip graft, using a straight root 2 to 3 inches long, and a scion 6 to 8 

 inches long. The grafts have been made at various times in the late 

 winter and early spring, most of them in February or early March. For 

 the most part they have been made by student amateurs, and yet they 

 have been as well made as the average of commercial work. It has ap- 

 peared that there is more dependent on the way the scions were handled 

 before and after grafting than in the skill with which the union was made. 

 To test the necessity for large contact of the cambium layers five different 

 methods or degrees of matching were tested, as follows : — 



(a) Matched on one side only, not at top or bottom. 



(6) Matched on both sides, not at top or bottom. 



(c) Matched at top, not at sides or bottom. 



(d) Matched at bottom, not at sides or top. 



(e) Perfectly matched all around. 

 The variety used was Baldwin. 



Where it was desired to avoid matching, the scion or root was cut away, 

 if necessary, to make a space of at leg-st 1 millimeter. The grafts were then 

 planted and cared for in the usual way. The results are shown in Table 1. 



