13 



bud generations. Typical fruits of some of these strains are shown in Plate 14. 

 These strains for the most part have been unintentionally propagated by nursery- 

 men and others. Some of them have been found to be very undesirable from 

 the economic standpoint, while others apparently possess some value for par- 

 ticular conditions of culture. 



Varieties of economic value and importance often degenerate or run out 

 through the unintentional or intentional propagation of undesirable strains origi- 

 nating as bud mutations. In the Washington Navel orange variety as grown in 

 California, the writer has found that the undesirable Australian strain, having 

 rank-growing trees, an example of which is shown in Plate 15, which character- 

 istically bear small crops of fruit of inferior quality, as shown in Plate 16, has in 

 some instances been propagated to such an extent as to render the variety an 

 unprofitable one. In a few cases the Australian strain has been intentionally 

 propagated by nurserymen because of the vigorous, hardy growth of the trees 

 and owing to a lack of knowledge of their fruiting habits. However, the propa- 

 gation of the Australian strain has usually been unintentional. The trees of this 

 strain produce unusual numbers of vigorous-growing branches, frequently called 

 suckers. This growth was formerly highly prized and commonly used for the 

 propagation of the Navel orange variety in California. Inasmuch as the supply 

 of such bud-wood was much more abundant in Australian trees than in the high- 

 yielding and desirable trees, budders naturally secured more bud-wood from the 

 Australian than from the Washington or best strain trees where it was obtained 

 without regard to fruit production. 



The isolation of the valuable strains of commercial varieties through bud 

 selection is one of the important factors in plant improvement work. After 

 strains have been isolated, their value for cultivation determined, and their char- 

 acteristics recognized, they are frequently classed as agricultural varieties. 



In potatoes, Myers (36) reports the results of hill selection from high-yield- 

 ing and low-yielding strains of potatoes in New York. The Cornell experiments 

 showed a five-years' average yield of three low-yielding strains of 82 bushels per 

 acre as compared with a five-years' average yield of three high-yielding strains 

 of 208 bushels per acre. The increase in yield secured through bud selection in 

 these experiments amounted to an average of 126 bushels per acre. Similar re- 

 sults were obtained by Daniel Dean (36) of Nichols, N. Y., who began hill selec- 

 tion with his variety of potatoes in 1904 and whose records show a difference in 

 production in his high-yielding as compared with his low-yielding strains of 280 

 bushels per acre in 1912 and 140 bushels per acre in 1913. The difference in 1913 

 would have been greater except for an early frost. G. D. Brill in one year's hill 

 selection work secured an increase of 107 bushels per acre as shown by the dif- 

 ferences in yields of a high-yielding as compared with a low-yielding strain of 

 the same commercial variety (36). 



The comparative yields of high- and low-yielding strains of the Carmen No. 3 

 potato variety isolated through hill selection during 1904, 1905, and 1906, are 

 reported by Waid (75) of the Ohio Agricultural Experiment Station. In these 

 experiments the high-yielding strain produced a three-years' average of 774 



