6 
GIOVANNA TOGLIATTI 
Position of points used for bridging 18-1 ; 18-6. 
<g> Position of points used for bridging 6-1 ; 6-3. 
-f- Position of points used for scale transfer 18-6. 
Position of points used for scale transfer 6-3. 
Fig. 1 
them, we measured the coordinates of 27 points placed in such a way that, on each 
model, 18 could be found in groups of three next to one another in the classical 
position ; 9 of these 18 points belong also to the subsequent model. The schema we 
have chosen has made it possible to perform the orientation using either 18 or 6 
points per model. For the scale transfer we used the points marked with a cross : 
according to the type of bridging, they were respectively 6, 3 or only 1 next to the 
nadiral point of the photogram on the left. This special choice of points was made 
because the theoretical research and the determination of the weight cofactors had 
led us to the conclusion that, if we put 12 or 18 orientation points in groups of 2 or 
3 very near one another, in the same position where we usually choose only 6, the 
weight cofactors decrease of 30 and 42 % respectively. Obviously, also the errors 
of the orientation parameters decrease of the same percentage. 
These results are true only in the hypothesis, certainly not verified in practice, 
that the relative orientation errors have but an accidental nature. If we take into 
account the influence of other factors, like for instance the residual distortion, it 
could very well be that a different distribution of the orientation points might be 
more convenient. We shall deal about this subject at the end of the present section. 
The results of the bridgings, that represent a remarkable amount of material, 
have been studied through the exam of the residual parallaxes in each model and 
of the deviations, that is the discrepancies between the coordinates of the 9 points 
placed in the overlapping zone between two consecutive models. The coordinates 
of these points are actually computed twice : the first one when we plot the model 
on the left, in function of the orientation elements proper of the central photogram 
of fig. 1 ; the second during the plotting of the model at the right, therefore intro 
ducing the orientation elements relative to the photogram which follows the pre-