pese ea 
  
  
  
  
the various values in this case reveals a few very high values, probably 
due to false stereo-effect because the shadow of the upper part of the 
head of the neighbouring tree coincides with the head of the tree in one 
of the prints. Difficulties of this kind are liable to occur when the direc- 
tion of the tree projection coincides with the direction of the shadows 
and the relative length of the shadows is great. 
The repeated measurement on model 15a with ocular magnification 
3x shows that it does not pay to magnify a print beyond a certain limit. 
If this limit is exceeded the print will loose its distinctness and the 
measurements become difficult. 
2. Comparative Measurements on the Scale of the Contact Copy 
Table 4. Comparisons on the Scale of the Contact Copies 1:20 000 and the Same 
Pictures Enlarged to 1:10 000 (Model 15a; Ocular Magnification 8x). 
  
  
  
  
  
| ET i TE EE o Rn | 
* | Scale 1:10 000 | Scale 1:20 000 | 
Interpreter | Systematic Standard | Systematic | Standard | 
error, m deviation, m | error, m | deviation, m | 
| | 
| 1 —24 | - 094 — 85 250554. 
| 2 — 0.8 | + 1.44 — 1.7 + 2.02 | 
| 3 "Hs | ii —06 on 
  
Table 4 shows that two of the operators obtained a greater systematic 
error in measuring on the scale of the contact copy than on the enlarged 
print, while the third interpreter arrived at the same result in both 
cases. As the factor by which the value obtained in the micrometer is 
to be multiplied in order to give the height of the tree is 20.5 for the 
enlarged print as-against 40.3 for the scale of the contact copy, eventual 
difficulties in adjusting the measuring mark ought to cause relatively 
great deviations in the values. However, the results do not indicate that 
measurements on the scale of the contact copy are so much more in- 
accurate than measurements on enlarged prints. The standard deviation 
obtained by interpreter 1 on the contrary indicates that the results in 
measuring on a smaller scale may be very even, and if further the mul- 
tiplication factor is taken into consideration, it is reasonable to assume 
that the contact.copy gives a better distinctness than the enlarged print. 
In expressing their general judgement all three interpreters found 
the contact copies very distinct and easy to measure on, as regards the 
ground as well as the tree tops. 
3. easurements on Negatives 
In order to make it possible to estimate the influence of the various 
instruments on the results, measurements have been performed in 
8