98 



promise. While the Rusli is the best American hazel that we have, 

 there is no reason to suppose it to be the best in existence, in fact it 

 seems most unlikely that it is and that the improvement already made, 

 great as it is, could probably be made much greater if we had better 

 material to work with. 



A nut arboretum is a place where is assembled all the best varieties 

 of each species of trees or shrubs which bear valuable nuts. It should 

 contain also trees which bear nuts outstanding in each particular qual- 

 ity, such as size, cracking quality, flavor, etc., even though they should 

 possess undesirable qualities which would prevent the nuts from rank- 

 ing high when all qualities are considered. It should contain also the 

 best varieties it is possible to obtain of all nut species related to those 

 which bear valuable nuts. For example in the case of the hickories it 

 should contain the best bitternuts, Carya cordiformis, the best shell- 

 barks, Carya laciniosa, the best mockernuts, Carya alba, etc., even 

 though we have no varieties of these species that bear really valuable 

 nuts. The bitternut is useful because it has thin shell and large pro- 

 portion of kernel and imparts these qualities to its hybrids. The shell- 

 bark bears nuts way ahead of any hickory species in size, and the mock- 

 ernut is the most rapid growing of the hickories, particularly on com- 

 paratively poor soils. The slow growth of the hickory is one of the 

 things that has so far prevented its attaining the importance tliat the 

 fine quality of our best hickory nuts would seem to warrant. To sum 

 up, a nut arboretum should contain species of all nut trees and shrubs 

 that seem at all likely to furnish material for breeding varieties better 

 than we now have. 



It may be asked why it is necessary to assemble all these varieties 

 and species in one place. Why, for example, could not a choice shag- 

 bark tree growing in Massachusetts be pollinated with pollen from a 

 superior bitternut tree growing in Iowa. While it may be possible to 

 do some work with widely separated parent trees, yet usually such 

 is very expensive and there is much wasted effort. Widely separated 

 trees also cannot be properly observed except by having an observer in 

 each place, which adds expense even if it can be effected, and this will 

 many times be found to be impracticable. The female parent trees in 

 particular should be conveniently located so that the growing hybrid 

 nuts may be properly marked during the summer and protected from 

 boys and squirrels during the fall. Those who have done plant breed- 



