acre necessary for wood production at this rate varies within wide 

 limits, as shown in the footnotes to Tables 64, 96, 128, and 160. 

 The plots measured as a basis for the yield tables in this study 

 were all normal, or fully stocked. They were selected in the 

 field because the number of trees per acre and their distribution 

 appeared satisfactory, and because there were no large openings 

 between the crowns of the trees. Final judgment on their nor- 

 mality or fullness of stocking was based on comparisons of their 

 total basal areas (sum of the cross-section areas of all trees at 

 breastheight) with the average of all similar plots measured. 

 Those found to be abnormal by this test (see the references cited 

 in footnotes 2 and 3) were eliminated from subsequent computa- 

 tions. 8 Because of frequent ground fires, hog damage, and lack 

 of sufficient seed trees, normal stands are comparatively rare 

 and represent but a fraction of the second-growth stands in the 

 South. In selecting plots some leeway as to the meaning of full 

 stocking was necessary in order that a sufficient number might 

 be found without unreasonable expense. Since the tables give 

 the average figures for the plots studied, they do not, strictly 

 speaking, represent maximum possible volume. The additional 

 fact that the great majority of the stands chosen had at one time 

 or another been burned over (few unburned areas are known) 

 and had developed naturally, should also assure higher yields 

 than are shown in the tables, when protection and management 

 are introduced. 



8 The elimination of abnormal plots explains the discrepancy in the number of plots 

 given as the basis for the height-growth classification graphs, Figures 1, 2, and 4, and 

 the number of plots listed in Tables 64, 96, and 160. 



48654—29 2 17 



