reference (shq), which we called "mean 
cylindric reference", or shortly "Coliseum 
reference" (fig. 1). 
  
fig. 1 
mean cylinder is an el- 
The reference 
liptic right vertical cylinder, whose 
plant is the first or the second of the a- 
bove said ellipses E1 and E2. 
The transformation of the local cada- 
strial coordinates  (ENH) into Coliseum 
coordinates (shq) is fully described in 
the Kyoto paper; we will not bring here 
the subject again. 
1.3 — 
phic frame for the whole monument, 
local surrounding triangulation net. 
This fundamental net consists of 8 
points, placed on sound pillars with auto- 
centering devices; the points were deter- 
mined with millimetric accuracy by conven- 
tional procedures (Table 1, Annex 1). 
Some other restricted  densification 
nets, ‘and the control points for the "ae- 
rial" triangulation on the walls, were de- 
rived frm this net. A sub-centimetric ac- 
curacy was obtained everywhere. 
The operations for the nets, and their 
results are described in the Kyoto paper. 
We ‘shall see later on, in the aerial 
triangulation adjustment (para 2.3), a re- 
liable check of the controls” 
We obtained a very good topogra- 
with a 
accuracy. 
1.4 - For the photogrammetric takings of 
the walls two strips were planned, a lower 
one with horizontal photographic axis, and 
D u 
521 
eJ 52300 
an upper one with inclined axis. We hoped 
to carry out horizontal takings of the hi- 
gher strip from an elevator carriage, and 
this we did at first; but unfortunately 
the carriage was no longer available, so 
that in the Northern upper strip we had 
only inclined takings. 
The taking plan, carefully studied by 
Prof.Carlucci, provided large longitudinal 
and trasversal overlaps, in order to ob- 
tain double alternate strips and a strong 
aerial triangulation block. The takings 
were effected at a mean distance of 30 m, 
with a P31 Wild camera, f= 100 mm, auto- 
diapositive colour film, photographic sca- 
le 1:300. Each taking was repeated twice, 
in the same position and attitude. 
2. - The "aerial" triangulation 
2.1 - Considering the big number of pho- 
togrammetric pairs necessary to obtain the 
complete coverage of the walls (about 
150), we deemed it opportune to determine 
their control by a particular kind of 
"aerial" triangulation.  I.e., ‘| dn’ ‘aerial 
triangulation in which the strips are the 
sets of terrestrial consecutive takings 
around the monument, as if they were taken 
by an aircraft who flew around it. 
So, both in El as in E2 we have a block 
of two strips. Each block has about 80 
photograms, with large overlaps and a lar- 
ge number of well determined control 
points. The fig. 2 shows their distribu- 
tion on the Northern El wall; in E2 it is 
repeated almost with the same standard. 
2.2 - At the beginnings we tried to ope- 
rate by independent models, taking into 
account the variations of the x distances 
due to the elliptic curvature of the refe- 
remce surface. But we were soon convinced 
that this was not a good approach, and we 
fell back on the projective bundle solu- 
tion. In fact this is the only procedure 
which permits to adjust the E1 or E2 chai- 
nings in an unique block, as it does not 
require an unique reference plane, and ac- 
cepts however inclined photograms. 
ed 2 
pm m a a o a a 5 \ 
  
angoanangaanganananenananananannnuo 
agaaanganaosocaosagnananaoaonan8a000000. 
/. ia a OD Cea D Cr C C8 Oe. D D DH. CQ CQ DH DELL À 
    
     
= a 
558 563 565 
$50 553 559 562 566 
  
fis. 2 
  
334