The survey underway of the dome of the SS. Annunziata has enabled 
us to better state this operative practice which at first glance does not 
seem to present great difficulties. 
In fact, once the station pattern for absolute orientation of the vertical 
stereograms has been arranged, it is necessary to establish their spatial 
position within the terrestrial topografic reference system X, Y, Z. 
This position can be easily stated with the instruments and methods 
of classic survey, where the points have been obtained in a stable way 
on the vault to be surveyed or they show up on the vault, in the case 
of painted details, very clearly. 
It is much more difficult to state the spatial position X, Y, Z of the 
control points when these result from a grid projection or geometric 
figures. In this case, the survey must of necessity use the method applied 
in « aerial triangulation » in order to state their spatial position. 
The geometric control to be used for obtaining blocks of vertical photo- 
grams will then be studied carefully, following the known practice 
evaluating in due time the limited possibilities offered by architectural 
structures such as those in question. 
Assumed that, as a basic working hypothesis, the varying lines and 
surfaces which constitute structurally the monument under survey be 
geometrically defined, it will then be best to measure the distances, 
the heights and the depths to be used instead of these control points 
having the known position X, Y, Z. With these geometric measure- 
ments, absolute orientation of the stereoscopic model will not result as 
being of great extrinsic accuracy, but certainly of sufficient accuracy for 
photogrammetric survey and later for planned researches. 
Photogrammetric plotting of the outside and inside exposures does 
not generally present great difficulties since the photogrammetric takings 
have been studies and accomplished with care. 
There is no doubt the final accuracy of a numerical or graphic survey, 
made for architectural purposes, depends always on the care which has 
been given to the stereoscopic takings. 
The choice of the plotting instrument must reflect, of course, the ne- 
cessity to scan, during plotting, optical models which are on a rather 
large scale, so that their graphic plotting is be done with magnification 
ratios which are not very large. 
Relative orientation of the photogram can be obtained either by a nu- 
meric proceeding or by an empiric optic-mechanic one. Where it is 
not possible to consider the 6 classic points — as occurred with the 
plotting of the inside and outside photographs of the dome of Santa 
Maria del Fiore(1) — it is always wise to replace the missing points 
with others strategically placed in the stereogram.