l86 I'LA.XT rKOl'AGATION 



234. Pedigreed trees again (see 232). J. P. Stewart discusses 

 this question in a recent bulletin, condensed as follows : 



Various theories of observation have been advanced, but only a 

 few experiments conducted. The net results of observation ami 

 discussion, however, have shown merely that important variations 

 do exist among mature apple trees, in almost any direction desired, 

 but thus far none of the variations within a variety, with the 

 apparent exception of color, have been actually proved to be herit- 

 able. In addition the most fundamental and generally accepted 

 theories are all against such inheritance, without excepting color. 



With this situation in view, a preliminary test was started at the 

 Pennsylvania Experiment Station on the influence of cion selection 

 in improving yields. The individual trees were chosen merely on 

 the recommendation of commercial growers, for the most part, 

 and without any definite and comparative records covering several 

 years — the only satisfactory basis for making the primary de- 

 termination of what are really superior individuals. This defect is 

 being remedied in another more comprehensive test recently started. 

 In the present test cions from supposedly superior individuals were 

 top grafted chiefly on Northern Spy stock. Ordinary nursery trees 

 of the same variety were planted alongside for comparison. 



It is much too early to draw conclusions from this work, but 

 the results to the close of the seventh year indicate that in four 

 of eight varieties, a slight superiority is shown by the trees pro- 

 duced from, the selected cions. In three varieties no advantage is 

 shown in either group, although the cions for these trees were 

 secured from one of the most prominent advocates of the practice: 

 in the eighth a slight difference favors the nursery trees. 



The results are evidently insufficient either to approve or con- 

 demn the practice of cion selection. Such trees may have more 

 certainty of trueness to name where the cions have some recently 

 from the pure trees of known bearing habits, and one is naturally 

 on the safe side of the question in using them, when they can be 

 obtained without material difference in price. This, however, is all 

 that can be said at present in fa\or of the practice so far as apples 

 are concerned. Much more data are needed. 



235. Stocks in top-grafting. — In case experiments 

 .should definitely prove cion selection to he advisahle. it 

 may be well to know something- of the relati\e values nf 

 various stocks available for toivo-raftiiio-. It seems also 

 that some desirable varieties shotihl reg^ularlv l)e to])- 

 grafted to secure better and healthier trunks. To secure 

 data on the relative merits of certain well-known varieties 

 for stock purposes, a test has been started at the Pennsv 1- 

 vania station, using- four trees each of three varieties to])- 

 grafted on five dift'erent stocks. Stewart's progress re- 



