14 
The principal tests carried out were: 
1) Reading by monocomparator TA 1/P-OMI of the photogram ste- ‘ a 
reo-pairs made by the TMK Zeiss, Contonar-Galileo and P/31 
Wild cameras. The plate co-ordinates of the 12 points obta 
ined with the facility and of about 50 points marked on the 
metal model were also surveyed, 
2) Reading by Stereocartograph 4th Off. Galileo of the model 
co-ordinates of these same points, 
The model was made using only the photogram stereo-pair 
made by the P/31 Wild camera. 
As regards the measurements done on the monocomparator, 
for all three cameras, the following orientations were done nu ft % 
merically: e 3 
a) a relative orientation on 6 points; an absolute orientation 
on the 12 control-points of the framework. 
b) a relative orientation on 12 points; an absolute orientation 
on the same 12 points, 
c) a relative orientation on 6 points (see (a)); an absolute 
orientation on 6 selected known points on the car model. 
As regards the measurements done on the Stereocartograph: 
a) an absolute numerical orientation on the 12 points of the 
framework; 
b) an absolute numerical orientation on the same 6 known points 6 e 
placed on the car model. 
The mean errors in the parallaxes, resulting from the tests 
done on the 6 and on the 12 points were very similar (mean er- . 
rors were about 0.10 mm in a completely casual way for the 6 
tests done); because of this the absolute orientations that we 
re compared were those obtained with a relative orientation on 
6 points. 
Obviously, the tests done while carrying out the absolute 
orientation using the 6 points placed on the car model allowed 
us immediately to determine the list of the spatial co-ordina 
tes of all the points marked on this model, in the co-ordina- é E 
sh