lal 
rh 
La ial WSFA 
al) 
ed 
LS 
The kite always flies downwind and within an arc of £509 from the wind 
direction. It is held by the navigator from two steering lines, which should 
be of the same length. A rough flying height estimation is obtained by knowing 
the length of the tether lines and the vertical angle to the kite. Changing the 
length of one line, i.e. pulling it gently, causes the kite to fly towards that 
direction. Learning to fly and steer the kite requires a few hours of practice 
and normal sensitivity to wind speed and direction. In Table 2 the relationship 
between wind speed and payload is given both for the DFM 2000 and the combination 
of DFM 2000 and Delta. Practical experience has shown, however, that it is safer 
to fly the kite, or kites, in constant wind speeds of 5 knots or more. 
3. Practical experiment 
  
Several considerations and previous field experience led to the choice of a 
suitable object for test photography. It should not present any major difficulties 
for launching and flying the kite and it should provide both visual and scientific 
interest. Height differences should also be present so as to provide a realistic 
test and evaluation. 
As & result of these considerations, the Uffington White Horse was chosen as 
the site of the experiment. It lies in Oxfordshire about 10km (6 miles) west of 
Wantage at the edge of the Lambourn Downs. The Horse itself is a large figure, 
somewhat abstract in appearance, but with well balanced proportions and cut in the 
thin grass layer that covers the chalk of the Ridgeway hills. The Horse is appro- 
ximately 150 m in length and 50 m in height and can only be seen in its entirety 
from about 2 km away in & northwesterly direction of from the air. The history of 
the White Horse is rather obscure.  Archaeologists seem unable to agree as to when 
exactly and why this, and other similar figures, were cut. The diverse opinions, 
however, appear to converge to the fact that this White Horse is one of the oldest 
examples, cut probably in the first century A.D. 
Considering its size, it was decided to photograph the front part of the figure 
only. Control points were established on the ground, in order to enable an assess- 
ment of the accuracy of the photogrammetric processing. These control points were 
evenly distributed and signalised with suitable targets. 
Horizontal and vertical angles to all the targets were observed from three 
theodolite stations in two rounds and two faces. A Zeiss (Jena) 010A theodolite 
(1") was used for the observations. Its major advantage was the automatic indexing 
which proved more convenient, since the height differences in the object were 
rather large. However, the difficulty of using such an instrument in windy con- 
ditions should be borne in mind. The slope distances between all three theodolite 
stations were measured several times, in both directions, with &n AGA Geodimeter 12. 
Simple arithmetic means of the reduced distances were taken to provide the scalar 
measurements. The triangle was solved as a closed independent traverse. This 
adjustment produced a mean misclosure of 16 mm in the horizontal co-ordinates. 
The control points were given co-ordinates in a local, arbitrary system by 
intersections. All possible combinations were used and finally arithmetic means 
were taken. An accuracy of 16 mm in X and Y was achieved. Trigonometric heighting 
was used for the Z values with a resulting accuracy of 25 mm. These accuracies 
were considered adequate as will be discussed later. 
The ground measurements were carried out at the same time as the photography. 
This ensured that the targets would not be moved and in this way a more permanent 
point marking became unnecessary. In this case the use of natural detail points 
was not possible, firstly, because they were poorly defined and, secondly, because 
access on the figure itself was not allowed due to potential erosion. The weather 
conditions were almost perfect. There was a steady and strong southeasterly wind 
throughout the day, together with a sunshine. Thus it was possible for two black 
and white (Ilford FPL, 125 ASA) and one colour transparency film (Kodak ,Kodachrome 
64 ASA) to be taken. A shutter spead of 1/250s was set on the camera, the aperture 
209