object is staked out on the ground, levelling on the ground has to be done in any 
case and the necessary corrections can be made then. 
Scale of Photographs and Maps. 
If the area concerned is covered by tropical jungle the scale of the aerial 
photographs need not be larger than 1:40,000. A scale larger than 1 : 20,000 is 
even undesirable, as the local difference in the heights of the trees would then 
disturb the overall impression of the stereoscopic image. By carrying out an aero- 
triangulation the height and position of an adequate number of points on each 
photograph can be determined in a system of co-ordinates. With the help of these 
points a basis can now be established for plotting the pairs of air photographs in 
a stereoplotting apparatus of the Wild A,, or similar, type. In this way a contour 
map is obtained, from each set of aerial photographs, of the area covered stereos- 
copically by those pairs. By carrying this out for each pair of pictures in succes- 
sion, a complete contour map of the region is finally obtained. The mapping-scale 
can be set at from 1:10,000 to 1:20,000. The distance between the contour 
lines is taken as being 20 m., but in some cases a distance of 10 m. can be taken. 
This distance is determined by the fact that all that can be seen in the stereo- 
plotting apparatus is the stereoscopic image of the upper surface of the trees. 
By measuring the height of the trees in open places or along rivers in each pair 
of pictures an estimate may be made of the average tree height, and this may 
be used as a correcting factor for the purpose of finally obtaining a contour 
map of the ground area itself. As, however, the heights of the trees very often vary 
rather suddenly, this method introduces a source of error which is, admittedly, 
unavoidable but which renders it impossible to achieve a greater degree of accuracy 
than is satisfied by plotting contour lines every ten or even twenty metres. 
Accuracy. 
Aerial surveys carried out in tropical regions for exploration and develop- 
ment purposes inclusive of road construction have already given us a certain 
amount of experience of the degree of accuracy to be expected. 
Figures 3 to 8 show a profile taken from the aerial survey map, and a pro- 
file obtained by surveying operations along the route of the road which was 
subsequently built, or from a map prepared by terrestrial methods. 
Fig. 2 shows the flying scheme of the area shown in Fig. 1, of which Fig. 3 
gives the comparison of the profile according to the photogrammetric map with 
that according to the terrestrial survey along the road after it had been built. 
The only ground control used for the elevation were the water levels at the 
coast and of the two rivers shown in the figure. 
Besides the errors caused by applying average corrections for tree heights as 
mentioned above, part of the discrepancies shown are due to the fact that there 
are small errors in the situation of the finished road as it was drawn in the photo- 
grammetric map in order to obtain the profile. This likewise occurred in drawing 
the profiles of figures 4 and 8, and in the case of figures 5 and 6, the profiles 
drawn in the photogrammetric maps may also contain small errors in situation, as 
it is difficult to find corresponding points on the photogrammetric map and the 
terrestrial map, since the latter has been compiled according to a local co-ordinate 
3