FRUIT BREEDING 275 



French prune tree which was first observed in 1904. Several 

 grafts from this branch were placed in bearing trees. These 

 grafts reproduced the characters of the sporting branch. In 

 1914, trees in alternate rows of an orchard were top-worked by 

 the use of buds from the new strain and compared with buds from 

 the normal French prune variety. The top-worked trees from the 

 bud sport bore larger fruit than those from the normal prune. 

 The fruits were also more evenly distributed over the tree than 

 in the original French prune variety. 



The above are some of the more striking instances of the pro- 

 duction of new varieties through the isolation of bud sports. 

 Crandall (1918) has made an extensive test in Illinois of the value 

 of bud selection in apples as a means of improving the variety. 

 Two distinct lines of study have been followed. 



1. The value for propagating purposes of buds selected in 

 different ways. The experiments included a comparison of large 

 versus small buds, of buds from different parts of the tree and 

 from different locations on the shoot. 



2. Selection of trees because of special merit. Comparison 

 of seedlings produced from large and small apples produced by 

 these selected trees. 



A considerable number of varieties was used for the first 

 study and a total of 5,400 buds were selected. A careful measure- 

 ment was then made of the yearly growth of wood from the buds 

 which had been previously selected. Growth curves were made 

 and on the basis of these results the conclusion was reached that 

 all buds from healthy shoots were of equal value for propagation 

 purposes. 



The characters of seedlings grown from seeds of large and 

 small fruits borne on trees of special merit were carefully studied. 

 Seeds from large fruits produced seedlings which were somewhat 

 more resistant to adverse conditions than seedlings grown from 

 small fruits. The hypothesis that this may be explained by the 

 fact that large fruits and large seeds frequently occur from crosses, 

 seems reasonable in the light of the work of Lewis and Vincent 

 previously cited. 



Stewart (1912) has discussed the value of cion selection in tree- 

 fruit improvement. Individual apple tree data over a period of 

 from ten to fourteen years were presented. Under apparently the 

 same conditions some trees were consistently higher yielders 

 than others. A review of considerable experimental evidence led 



