growing stock for instance. The estimation were done on en- 
larged, brilliant pictures on the scale 1:15 000 (4.2 inches to 
a mile). 
2. RESULT OF THE MEASUREMENT. 
2. l. Sources of error in measuring with stereometer. 
The horizontal parallaxes were measured by means of mirror 
stereoscopes and stereometers and from the parallax differences 
the elevation differences were computed according to the for- 
h ; S : ; 
mula 4 h= 4 p!. b? where 4 h is the difference in elevation, 
) 
| pt is the parallax difference, h is the flying altitude and b! 
is the photography base measured in millimetres on the photo- 
graph. 4p is measured by means of the stereometer, h is cal- 
culated from the photo scale and the focal length (c) of the 
camera and b! is the distance between the principal points of 
the photographs, measured on one of the pictures. Provided that 
the terrain is plane and horizontal and the pictures are taken 
with the camera axis absolute vertical and the photography base 
horizontal, the formula will theoretically give accurate height 
differences (except for errors caused by the fact that the for- 
mula is approximate, see below). However, in praxis one has 
to take into account approximations in the respects mentioned 
above. To be able to judge the results of the measurements one 
has to get an idea of the magnitude of these errors and their 
effect on the determination of the height differences. 
2. 11. Determination of the flying altitude. 
The information about the flying altitude given by the altitude 
barometer of the air craft is normally not sufficiently accurate. 
For more precise measurements it is wise to check the altitude 
by means of scale examinations on the photos. A scale deter- 
mination requires a base known in the terrain and possible to 
identify in the picture. Normally one or, preferably, several 
such bases can be obtained from some kind of a map. Care 
should be taken that the ends of the base lie on the same level. 
Thus the flying altitude corresponding to this level will be 
obtained. By stereoscopial viewing of the pictures it is easy to 
J