Treatment 
In-preparing the statistical data the heights measured in the field 
have been used for comparison. The mean value of the errors has been 
regarded as systematic error in measuring. After every error had been 
corrected with regard to the systematic error the standard deviation 
was computed. 
The systematic error thus includes certain errors which are not syste- 
matic but cannot in this case be separately considered. 
Results Obtained in Measuring Tree Heights 
The experiment has been planned in such a way that many of the 
‘sources of error described in the previous survey [1] may be expected 
to have been eliminated, The remaining errors may — if not entirely at 
least chiefly — be assigned to the operator, the possibility of identi- 
fying the top and the root of the tree, and the instrument. To what 
extent these factors may have influenced the result will be discussed in 
connection with the various measurements. 
i. Measurement on Paper Prints, Scale 1:10 000 
As table 1 shows the photographs which have been used for the 
investigation were taken some in cloudy (prints without shadows) and 
some in clear weather (prints with shadows). The pictures were select- 
ed with a view to obtaining greatest possible variations of illumination. 
All pictures were on examination found to be of excellent quality. 
Table 2. Systematic Errors and Standard Deviation in Measuring on Paper Prints, 
Scale 1:10000 (Ocular Magnification 8x) 
A bla A — 1280 m/31 A — 1580 m (410 À — 1890 m 15a A — 3120 m 
| Model | | | 
  
  
  
  
  
No shadows | No shadows | Shadows | Shadows | 
| mue ^ pr cs ea empresas | crests renee ét - 
| | | 
System) Stand. | System. | Stand. | System. | Stand. | System | Stand. | 
[nterpreter | | | | | | | 
| error | dev. | error | dev. | error | dev. | error | dev. | 
| | | | | 
1 | —5.66|-2:188| —48 |—c 1921: 1210671 — 2.4 | + 0.99 
9 | B64 1.61 20 {41.541 400% 1.50) —08 | + 1.44 
3 — 96 | + 1.47 | —08 | +230] —06 | + 1.54} —01 | + 1.00 | 
  
Ne NS ri MÓMÁÀM 
  
A comparison between the systematic errors in table 2 shows a strong 
tendency to underestimate the tree heights on the prints without 
shadows. It is also interesting to notice the great difference between 
operator 1 and operators 2 and 3 as regards systematic errors. 
So as to make it possible to form an idea of the difficulties encount- 
ered in measuring on various models, the operators in recording the 
results were to make a note concerning the factors which had caused 
4