1917] Bruce: A New Dendrometer 61 



The instrument shown has an arm eighteen inches long and will 

 read diameters from three to seventeen inches. A longer arm is, of 

 course, possible, but at about thirty inches a point is reached at which 

 the adjustment of the sliding mirror when held in working position 

 would become awkward. This may then be taken as the practical 

 limit, unless some modification of the principle be adopted. This 

 range will be sufficient for a great deal of the work to be done. If 

 less accuracy is required, measurements of double this size can be 

 secured by taking the center of the tree as the target for the direct 

 line of sight, instead of the left-hand edge, and bringing the reflection 

 of the right-hand edge in line with the center point. This operation 

 can be performed more accurately than might at first be supposed, 

 and the method, while rough, is probably quite adequate for work in 

 connection with the Pacific coast volume tables already mentioned, 

 in which taper is a factor. 



A quick field test of the parallelism of the sight lines consists in 

 measuring the same diameter at two different distances. The readings 

 should, of course, be identical, or rather, since a small observational 

 error is unavoidable, as nearly identical as would be two consecutive 

 measurements from a single position. If an error is found and it is 

 not convenient to make the proper adjustment it may be simply and 

 quite accurately allowed for, by taking consecutive observations at 

 two known distances. For example, suppose the first reading is 14.8 

 inches and the second reading taken at one-half the distance is found 

 to be 14.4 inches. Since the error is proportional to the distance, a 

 reduction of the distance to one-half must also reduce the error to 

 one-half. The reduction in error is .4 of an inch, the total original 

 error must have been .8 of an inch, and the correct reading is therefore 

 14.8 — .8 = 14.0 inches. Where the errors are small, the major por- 

 tion of them can thus be eliminated, even if the distances are estimated 

 instead of measured. 



A modification of this type of dendrometer is suggested for timber 

 survey crews which are using volume tables to a fixed top-cutting limit 

 such as six or eight inches. All that is necessary in such cases is the 

 pair of parallel mirrors, one of which is adjustable, mounted six or 

 eight inches apart on any light but rigid base not affected too readily 

 by changes of temperature or humidity. By thus eliminating the 

 straight edge and slide of the instrument herein described, the most 

 serious source of error will be eliminated and the cost largely reduced. 



