to near vertical as well as to convergent photography. This method, also programmed by
D. Eckhart, has already given a proof of its high quality. It will be published, probably
by the end of 1960, in the thesis which Ir. v. d. Hout will present for his Doctor’s Degree
to the Delft Technological University.
Concluding this series of contributions of Delft to the improvement of aerial triangula-
tion we should not forget a moderate instrumental contribution which nevertheless is of
importance: a new device for stereoscopic point transfer. With the more frequent applica-
tion of block adjustment it is obvious that the identification of tie-points in adjacent strips
needs a high precision, since errors in identification will increase the residual errors in
these tie-points to their full amount after the block adjustment. The first method in use is
to select natural points. In many types of terrain, such as forest and jungle, this is either
barely or entirely impossible. In more cultivated terrain we can identify natural points,
but many easy points give only a very low precision. For these points mean square errors
in transfer between 20 and 40 microns can be expected. These values are much too high,
compared with the values which can be obtained as relative errors after block adjustment.
In some kinds of terrain it is possible to select very small black dots of only a few microns.
Their precision will be sufficient but to find these in the always small areas of the overlaps
takes much time and these dots require special indication in order to avoid confusion.
The solution for the marking of points could be obtained in Delft along two lines, both
already in use there since 1935. The first was applied in the radial triangulator of De
Koningh (see National Report of 1938 in International Archives of Photogrammetry
Volume III, fase III). In this the negatives are observed in the normal way as in the radial
triangulator but with emulsion down, and after the floating mark is set on the ground,
small holes are drilled with a special drill which comes up from below. This solution has
been accepted by Wild in its new P.U.G. The I. T.C. has preferred the improvement of
the other Delft solution of 1935 (see the same National Report 1938) using a free hand
motion of small transporters with floating marks, observed under a normal mirror stereo-
scope with Dove-prisms in its telescopes. The floating marks were little holes, through
which a hole is pricked with a needle in the emulsion at the moment that the floating
marks are set on the ground. This free hand method gives a mean square error in the
transfer of points between 15 and 30 microns, depending on the training of the operators
and the material. P. Dongelmans, chief mechanic of the I. T.C., improved these simple
devices by replacing the holes in the transparent material of the transporters by very small
steel spheres pressed into perspex plates. These steel spheres act as floating mark and after
setting these marks by means of micrometer screws instead of by free hand, they are beaten
from above with a constant force. The spheres press a small hole in the emulsion. The
mean square error in the transfer of points obtained with the use of these new trans-
porters and normal mirror stereoscopes is so far 8 to 12 microns, also depending on the
skill of the operator and the quality of the photographs. Patents have been applied for
these transporters in several countries and Zeiss Aerotopograph will bring these small
instruments on the market. They will be shown at the Zeiss exhibit at the London Congress.
III. 2. General remarks about the execution of aerial triangulations
Regarding the execution of aerial triangulations by the various services in the Nether-
lands we first make a few general remarks in common for about all of them.
There has been about no calibration at regular intervals of the cameras. In general
the data provided by the factories was trusted. As mentioned in cHAPTER I, I considered
this as a weakness of the system, in particular with the use of the Zeiss twin camera.
12