LUTHER BURBANK 



be made unless the fruits are constantly tested in 

 order to select the best for the continuance of the 

 experiment. And as there is no known substitute 

 for the human palate in making such selection, 

 the tasting of fruits must be regarded as an un- 

 avoidable part of the plant developer's every day 

 work. 



In the case of the Goumi Berry, my efforts at 

 selective breeding have been rewarded by the 

 notable progress of the plant, first in the elimina- 

 tion of the thorns, and secondly, in the improve- 

 ment of the fruit. 



Here and there I have found a seedling, the 

 fruit of which is pleasant to the taste, and by selec- 

 tion through successive generations a variety of 

 Elaeagnus has been produced that gives great 

 promise of eventually growing a fruit of real 

 value. 



My experience with the genus has included 

 tests of five species, all imported from the Orient, 

 bearing the specific names of Elaeagnus augusti- 

 folia, E. umbellata, E. pungens, and E. argentea, 

 in addition to the original E. longipes. There are 

 three closely related plants also belonging to the 

 Oleaster natives of North America, these being, 

 E. canadensis (sometimes called Shepherdia cana- 

 densis), E. argentea, the buffalo berry (sometimes 

 called Shepherdia argentea), and E. argentea, the 



[248] 



