REPRODUCTION OF WESTERN YELLOW PINE. 



77 



merit station show that after the first winter the greatest losses are 

 due to drought in the latter half of June and the first half of July. 

 Shading does not enter, because all excessively shaded plots have been 

 eliminated from these records. Comparing the group-selection and 

 the scattered seed-tree methods, the consistent superiority of the 

 latter might be attributed to the elimination of root competition due 

 to heavier cutting. But this theory is controverted by the shelter- 

 wood area, which during the first period showed the highest survival 

 of all. Again, if we attribute the high survival in the shelterwood 

 cutting during the first period to protection, we must account for the 

 fact that survival in the scattered seed-tree cutting exceeds that of the 

 group selection, and during the second period even surpasses the 

 shelter wood. 



Further inquiry into the development of seedlings and the char- 

 acter of sites throws additional light on the problem. In the ex- 

 aminations of September, 1921, all 1919 seedlings were classified as 

 to relative size and vigor. Unusually large and thrifty seedlings 

 were classed as " very large." They are generally over 4 inches in 

 height, measured to the terminal bud, with proportional thickness 

 of stem and length of foliage. Somewhat smaller but still vigorous 

 looking specimens were classed as ;i large." Those which were of 

 only mediocre development, usually not over 2 inches high, were 

 classed as i; medium " : and the very smallest were classed as " small." 

 (PL XII. Fig. 2.) The results of this classification are shown in 

 Table 20. 



Table 20. — Effect of different methods of cutting on size of seedlings. 





Basis, 

 numoer 



of 

 •eedlings. 



Percent. 



m Method of cutting. 



Very 

 large. 



Large. 



Medium. 



Small. 



G r oup selection. . 



48 

 139 

 301 



6 



20 



1 



34 



23 

 8 



38 

 40 

 53 



22 



Scattered seed tree 



17 



Shelterwood 



38 



A glance at the above table shows that the shelterwood method 

 ranks far below the other two with respect to size of seedlings, and 

 that the scattered seed-tree method ranks highest in percentage of 

 very large seedlings. These extraordinary specimens occur almost 

 exclusively on stump patches. Stump patches occupy the greatest 

 proportion of area in the scattered seed-tree division and the least 

 in the shelterwood. This fact explains why the scattered seed-tree 

 method leads in both survival and rate of growth. That the group- 

 selection method should excel the shelterwood method with respect 

 to size of seedlings, is obvious, but why it should lag behind in 



