ground control was obtained by means of normal geodetic survey. With 4 of these blocks 
the adjustment was carried out by means of the pre-war Schermerhorn method and VI and 
VII by means of the I. T.C.-Jerie analogue computer. We draw the attention to block 
number VI which is of a size that could be usual in cadastral surveys and gives a very 
high relative accuracy of 13 microns in the scale of the negative and 20 microns as the 
mean square error in the absolute position of the tie-points. What is less satisfactory in 
this case is the long time per model of about 3 hours which makes 6 hours per section. 
This is longer than any other block adjustment according to this method. This, however, 
is partly due to the larger number of points in each model, which requires more time for 
transformation. This can perhaps also explain the 76 minutes for bridging of a model, 
compared with the 62 minutes for block VII, notwithstanding the 9 points used for 
numerical orientation in VII and only 6 points in block VI. The influence of the content 
of each section is even more evident in the blocks I a and I b. Each section took there 
17 hours. This makes only 58 minutes per model since a section consists of 18 models. 
The 17 hours are not used for the layout and its adjustment but firstly for the formation 
of the sections by linear transformation of the 18 models on each other and for the deter- 
minations of the final coordinates of all points in each section by lineair transformation, 
using the definite transformation formulae, resulting from the block adjustment. K.L.M. 
Aerocarto states in the Netherlands report to Commission III that the use of aerial trian- 
gulation is increasing since the analogue computer of Dr. Jerie was introduced and that in 
future most orders will be carried out with the aid of aerial triangulation. In another 
report this organization mentions that the Jerie method gives excellent results but that 
it must be considered a disadvantage that the ground control points must be located at 
special places of the area, namely at the corners and along the perimeters. I might add 
to this that it is very unlikely that any method can give reliable results without ground 
control at the extreme edges of a block. Otherwise extrapolation takes place and in loose 
corners the errors are entirely uncontrolable. The edges are the normal places for control, 
whether we have a block adjustment according to the normal slotted templet method 
or with a block adjustment according to the method of least squares. The only remarkable 
fact for the latter method of block adjustment is that additional control somewhere in 
the middle of the block or distributed over the whole area inside the block does not improve 
the results very much. 
III. 6. Aerial triangulation in the Military Topographic Service 
This Service carried out only radial triangulation of 68 strips with a total length of 
1180 km. Almost all flights were done in East-West direction by using the Wild Camera 
RC 5a-21 cm. Scale of photography about 1 : 20.000; longitudinal overlap is 55 to 70% 
and the lateral overlap is + 25%,. This service started to use the new radial triangulator 
Roelofs Wild RT 1 and with this instrument the production was 8 photopairs per 5 hours. 
After adjustment of the rhomboids and the calculation of the strip coordinates the strips 
were corrected for scale and azimuth by means of a second order curve, this curve being 
derived individually for each strip from the 4 superfluous ground control points given 
in every strip. After these corrections the average is taken of the two values derived in 
each adjacent strip for the common points. It is stated that the coordinates determined in 
this way would have a mean square value of the residual differences in the common points 
of adjacent strips of 70 to 90 cm. About 400 ground control points were given in the total 
area of 3000 square km. This, however, was not one large block but distributed over a 
great number of isolated areas scattered over the country. 
20