8 Forestry Quarterly. 



struction of the tables, as several consulting foresters have found 

 to be the case, an effective means of frustrating the work is to 

 delay sending numbered logs to the mill. This difficulty can be 

 avoided if the contents of trees measured in the woods are com- 

 puted from a log scale or 'tally' made at the mill. Errors in- 

 herent in this procedure are in most cases not excessive, par- 

 ticularly if butt logs are separated into a class by themselves. 

 Variations in 'defect' and method of logging, milling, and utiliza- 

 tion have so much weight in the case of hardwoods that the usual 

 fluctuation in the taper of the upper logs of a tree are for prac- 

 tical purposes negligible. In other words the process of follow- 

 ing numbered logs from the woods to the mill is a refinement 

 that is not justified by the accuracy of other phases of the work 

 and of the final application of the tables. 



A second procedure which deserves attention is the elimination 

 from volume table data of all defective and poorly shaped 

 trees. Graves in his 'Forest Mensuration' states, 



'Care is required in the selection of the trees for measurement. It 

 is the rule to measure only sound trees, because volume tables show 

 the full contents of sound trees. It might appear that the tables would 

 be more practical if based on average trees, including those partially 

 defective. But a table made up in this waj' would be extremely unre- 

 liable, for it is well known that the defects of trees differ greatly in 

 different situations ; so that a table based partly on defective trees would 

 be useless in eliminating trees whose defects are different from those 

 of the trees observed in the construction. Again, any such defect as in- 

 jury by fire, insects, disease, wind, or ice would entirely vitiate a table 

 constructed for trees showing another defect than the particular one in 

 question. Whereas a table based on sound trees may be reduced for un- 

 soundness in logs.' 



In most portions of the Northeastern United States a large 

 percentage of the hardwood trees are defective or deformed. 

 To sort out and measure only sound, straight logs and trees is 

 laborious and expensive. Furthermore, if tables are based upon 

 selected trees they must be discounted whenever they are used 

 even if tlie limber and methods of utilization resemble closely those 

 where the tables' were compiled. In addition the construction 

 and use of tables based upon sound trees presupposes an ac- 

 curate knowledge of the exact effect which different types of 

 abnormalities have upon volume and grade. Unfortunately these 

 converting factors are not available at present, and are difficult 

 to analyze because the influence of a given abnormality is not a 

 constant quantity, but fluctuates widely with variations in meth- 



