14 JOURNAL OF FORESTRY 



are therefore computed from the tables for the present time only. 

 Since the volumes at the present time, given in the tables, are for 

 dominant trees they will overrun somewhat the volumes of intermedi- 

 ate and, especially, subordinate trees, but this is believed unimportant 

 as the aggregate volume of these trees will be relatively small. After 

 eliminating the smaller trees which may be expected to die or be 

 removed before the end of the forecast period, the volume of the 

 remainder may be considered the same at the end of the period as at 

 present, or a slight increment may be added. 



It is not believed that forecasts on the above basis can be made with 

 any assurance for more than four or five decades, since the ability of 

 trees now dominant to remain for any long period even in a stand at 

 present fairly open, can only be guessed at. Two things are certain : 

 the shorter the period the more accurate the forecast ; and the younger 

 the stand the more nearly it will approach even-aged conditions and 

 the more closely will correspond the yields in the tables for even-aged 

 stands. Forecasts thus made ought to be checked by what is known 

 regarding the basal area, size, and number of trees per acre in mature 

 dense stands for the site under considerations. This method of growth 

 forecast is, of course, very approximate. Like the site classification. 

 it is proposed as a means of getting something at once which we could 

 hardly hope to obtain otherwise for a long time. As rapidly as data 

 become available whereby more accurate and refined results are assured, 

 the growth forecast methods here described may be laid on the shelf. 



