Periodical Literature 763 



classifying diameter classes or to confining the investigation to 

 certain stands and overlooking that shaft form varies from stand 

 to stand. 



Comparing with voltime determinations by 2 w sections, Flury 

 found the following deviations by employing the Huber formula. 



Sawlogs Building Timber Timber-wood 



Average of Top Dia. 12 Inch Top Dia. 10 Inch Top Dia. 3 Inch 



576 Spruces —3.6% —2.4% — .5% 



158 Firs —2.3 — .6% —3.1% 



479 Beech —1.6 —3.2% 



The errors are still greater, if as Flury has done (in Mitteilungen 

 der schweizerischen forstlichen Versuchsanstalt, vol. XI, 1916) 

 reference is made to tree diameter classes. Thus in spruce of 

 24 inch d.b.h. with a top diameter of 17 inch, the error percent 

 is as high as — 8.5, scaling down with a 3 inch top diameter to 

 — 1.8 per cent. 



Burger adds a less extensive series of measurements, relating it 

 to varying log lengths, and formulates the results of these various 

 investigations as follows : 



1. The cubing of middle diameter and length almost always 

 furnishes too small (negative) results ; 



2. The error percent decreases with increasing length; 



3. Positive differences can occur ; 



4. The deviations from the true volume are greatest (in spruce) 

 with a log length of 6 to 8 m from the stimip ; 



5. A remarkable fact is that the maximum of the error curve (not 

 the maximum deviation) occurs not with the timberwood diameter 

 (3 inch), but a few meters lower. 



The explanation of this last point is to be found in the fact that 

 the influence of the diameter, especially in long pieces, is of more 

 import than the lengths, especially when the diameter falls into 

 the portion of the tree, where there is a rapid diameter decrease, 

 when one centimeter decrease offsets 2-3 m in length. 



The author modifies two statements in Udo MuUer's Holzmess- 

 kunde. Muller states that "the accuracy of Ruber's formula 

 increases, the smaller the length of the log compared with the 

 length of the whole tree, so that 3-5 m logs are measured very 

 accurately, the more so the more they belong to the middle of the 

 tree." This does not hold good when the region of the root collar 

 is included. 



