77() JOURNAL OF FORHSTKY 



foot. In case the observer stands more than lOO feet away from the 

 tree the deviation from the true height is the product of error (in in- 

 correct interpolation) and i /lOO of the excess distance (over lOO feet) 

 from which point the reading is taken. An example will make this 

 matter clear. Suppose a reading taken at 200 feet records 63 feet, the 

 true height of the tree would have computed 124 feet — a difference 



200 



of I X = 2 feet. It is obvious that the degree of error becomes 



100 



greater with increase in distance (over 100 feet). The chance for 

 error under this condition will, however, be somewhat offset, l)ccause 

 actual readings (on hypsometer scale) are reduced in direct proportion 

 with increase of distances from the tree, and therefore, since the inter- 

 vals between the graduations on the hypsometer scale increase in pro- 

 portion to the lowering of the readings, closer interpolation to the 

 nearest foot is also possible. 



Where readings are taken less than 100 feet away the degree of 

 error will be afifected by an opposite set of associated factors. In that 

 case the effect of incorrect distances is small and the error decreases in' 

 direct proportion with the shortening of the horizontal measurements. 

 On the other hand, the intervals between graduations on the hypsometer 

 scale increase as the horizontal distances are reduced, and accurate in- 

 terpolation then becomes difficult. 



Considering each set of factors which may cause error in computing 

 true heights, it appears that the former are less effective than the 

 latter. This deduction is based on the fact that closer interpolation 

 to the nearest foot is possible when the graduations on the scale are 

 relatively wide.^ Furthermore, a better line of sight is obtained if 

 observations are made at relatively long distances from the tree. This 

 applies particularly where tall trees are involved, whose crowns are 

 often so flat that the true tip is difficult to perceive. It is then neces- 

 sary for the observer to choose a point sufficiently far away in order 

 that the angle of sight may be low. In most instances this distance 

 will be considerably over lOO feet. 



"In connection with this matter it is recommended that the scale of the 

 hypsometer be graduated to 2]/^ feet intervals, so far as possible. If this could 

 be done the 2^/2 marks should be scribed in red,- as this would then clearly 

 distinguish from the black 5 etchings. 



