Amongst other things, we agreed on the need that the various controls 
of the cj, (p, * rotations should always be turned in the same direction, 
to avoid the damaging effects of any back-lash. 
4. The actual bridging tests were performed on an aerial strip which 
EIRA had already examined for the analogical aerial triangulation ex 
periments carried out some time ago for Commission A of O.E.E.P.E., 
and belonging to the experimental polygon set up in Italy by the said 
Commission in 1954. In particular we examined a series of 21 photo 
graphs of the strip marked 5 A, taken at an average relative flight height 
of 3970 m and a ratio b/H = 0,5 with a plate camera SOM 125, size 
19 X 19 sq. cm (6). 
The tests of analogical aerial triangulation performed with the said 
material were four in all: for tests A and B we employed the control 
device, C and D were carried out without. They were performed ac 
cording to the method used for free bridging, which had already been 
experimented with the precision grids, suitably amending the effects of 
the earth curvature and of air refraction (7). The scale of the model 
was in all cases of 1: 10 000. 
All bridgings were carried out in the same direction of the strip 
and the points observed — both the base and the control points — 
resulted to be always the same. Furthermore, to eliminate any possible 
systematic effect of an initial nature special care was taken to find out 
each time the best absolute orientation of the starting stereogram. 
In order of time we first effected the instrument bridgings for which 
the control device was employed, and then the others. Nothing of special 
importance is to be noted as to the actual performance with the instru 
ment. They proceeded regularly in stabilized ambient conditions. 
After the instrument operations were completed we reduced the 
various coordinates to the same instrumental reference system, that 
relating to each first model. Then we transformed the coordinates, thus 
reduced, from the instrumental system to the terrestrial topographic 
system in the Gauss-Boaga projection, and we calculated the different 
unadjusted errors on the known control points. 
The results obtained in these four tests were graphically represented, 
to allow a better evaluation, on plates Nos. 2, 3, 4 and 5. On each 
of these we showed, with the adjustment sign, the differences A X, 
7