device are suggested. If calipers are not available j a diameter tape 

 can often be placed around the end of a log to measure its diameter. 

 Double bark thickness must be deducted. Unless the scaler has had 

 considerable experience, large errors will be introduced by measuring 

 across the tops of logs with a scale rule. 



Further complications in getting a good check scale come from the 

 difficulty of finding full-length trees to scale when the sale areas 

 are inspected. Modern hot logging methods do not leave an accumulation 

 of dorni timber in the vroodSo With sampling involved, this difficulty 

 will be even more pronounced, as only sample trees can be check scaled. 

 To facilitate finding the sample trees, at the time of m.arking and 

 measuring distinguishing marks should be placed on these trees, such as 

 ax blazes on three sides or extra-large paint streaks if the trees are 

 being marked with spray guns. Furthermore, check scaling must be re- 

 presentative of the timber. Samples must be obtained by checking a 

 few trees in scattered locations frequently, rather than by checking 

 all in one small area at one time. 



If the original sample is stratified, the check scale usually should 

 be stratified in the same xvay, that is, a representative check scale 

 obtained in each size stratum. If the correction factors found in 

 the various strata do not differ materially from each other the check 

 scaled data can be pooled for a common correction factor. The suita- 

 bility of a common correction factor can be tested by plotting all 

 the scaled volumes over their corresponding tree measured volumes. 

 Two main requirem.ents should be met: first, the relationship should be 

 linear indicating that the correction is the same throughout the range 

 of volume; and second, the straight line relationship should pass 

 through the zero point on the graph indicating that there is no constant 

 correction in addition to the percentage correction that must be applied. 

 The line passing through the zero origin also insures that small trees 

 do not have a plus correction and large trees a minus correction or 

 vice versa. 



SIZE OF SALIPLES 



The relative cost of tree measuring and check scaling affects the size 

 of the check scale sample required as well as the size of the first 

 sample. The final volume estimate is determined by two samples; the 

 first made in the standing trees and the second in the check scale. 

 If one is highly accurate, the other can be less accurate. The best 

 balance between the two must necessarily be based on the cost of 

 getting them. Only estimates are available, because costs have not 

 been studied systematically, but obviously, check scaling a single 



-20- 



