454 



Forestry Quarterly 



tion will depend on the object of the study. In studying even- 

 aged stands, Mr. Stetson simply ruled out the diameter classes 

 which would die before the expiration of the period for which 

 growth is being predicted. In uneven-aged, unculled stands Mr. 

 Stetson classified his trees according to the degree of suppression 

 as indicated by crowns. For such stands it is probably the best 

 classification to us 3, but in stands which were culled of their 

 spruce and best hardwood, 15 to 30 years ago, this kind of crown 

 classification will not produce good results. The writer found it 

 almost impossible to classify the trees in the field by this method; 

 moreover, the data, when computed, showed that there was no 

 real need for a such a classification. In these culled old growth 

 stands we are now using a classification for both the estimate and 



growth based on form; this includes the size and shape of crown, 

 total height, ratio between total and merchantable height, form 

 of bole and quality of lumber it will saw out. There are three of 

 these form classes and they correspond to the typical tree of the 

 three sites. Our classification grew out of the fact that in esti- 

 mating by the strip method in these culled stands so many typical 

 third quality form trees were found on second quality sites that 

 the site as a unit, over which one height curve could be applied, 

 was abandoned. Site was, however, considered as a basis for 

 dividing the forests into stands or sub-compartments. Instead of 

 making a separate height curve for each type, one was made for 

 each form class. The trees belonging to different classes were 

 tallied separately regardless of the type or site on which they were 

 found. Thus, the taking of height data was the first step in 

 making the estimate of any particular forest. The height data 

 and the knowledge of the timber obtained while collecting it 



