ut for 
ength, 
'eding 
t; the 
ne. 
with 
ziven; 
| com- 
from 
| first 
Z, for 
er re- 
were 
on to 
round 
)ssible 
strip, 
7). Of 
thods. 
arate- 
were 
'urves 
bases 
itlines 
s con- 
vident 
X f the 
those 
error 
carry- 
rated 
have 
s The 
OEEPE, COMMISSION A, SOLAINI AND TROMBETTI 31 
The maximum values of the gross errors in X, Y and Z have seldom been (15%, 5%, 
30% respectively) lower than 20 meters, and while those in X and Y have almost always 
been between 20 and 80 meters; those in Z have been much smaller (30%). 
These percentages, as well as the accurate comparative examination of the curves, 
seem to indicate that the method of bridging by bz — 0, also when operating with work- 
ing and ambiental conditions the most possible constant, causes errors having a slightly 
different law for the two different initial signs of the base. Finally the accuracy in Z 
looks inferior that that in X and in Y, that is the bridging by.ba= 0 looks as it cannot 
control sufficiently the errors in Z. 
3.2. The results obtained from the statoscopic bridgings are slightly different from 
the previous ones. The estimation of the resemblance in the two passages starting from 
internal and external bases has brought a better result (80%) due surely to the contri- 
bution of passages in Z conducted by the statoscopic data. And even better is the conclu- 
sion that can be drawn from the regularity (7096) of errors propagated along the strip. 
On the contrary, no real and substantial] improvement took place in the distribution 
of the maximum errors in X and Y: always few errors below 20 meters (15% and 10%) 
but again few greater than 80 meters so that the distribution remained concentrated 
between 20 and 80 meters (80% and 90%). 
On the contrary, the maximum errors in Z have profitted by the statoscope being 
very often (80%) within the limit of 20 meters. 
3.3. The solar bridgings have slightly different results as the statoscopic ones con- 
cerning resemblance (70%) between the two starts with bases of contrary and regularity 
of error propagation, (80%). 
On the contrary we can observe an improvement concerning the maximum errors in 
Y, more often (70%) contained between 20 meters than between 20 and 80 meters (30%) 
and which have in no case exceeded 80 meters. A more decisive step forward has been 
obtained from the maximum errors in Z which never exceeded 20 meters. Instead the 
solar data have not improved the distribution of the maximum errors in X which rarely 
have been within 20 meters (5%) and more often between 20 and 80 meters (65%). 
3.4. From a comparative point of view we noticed that in the bridgings with the three 
methods, the anomalies still took place almost always in the same points. These look as 
if they were due to the same defects of the photographic material; in fact the Centers 
carried out their work with diapositives taken out by inversion from the same original 
diapositive by serial processes. It is symptomatic, for instance, that for the three strips 
1A’, 1 A", and 1 A" taken out from the same strip having a very strong longitudinal 
covering, all the curves of error in Y deduced from bridging with the methods bz — 0 
and the statoscopic ones (there were not exposures by solar periscope) are decidedly 
negative. Only a longitudinal curvature of the negative film, due perhaps to calendering, 
can explain this. 
The conclusion which could be drawn is that methods of bridging which make use of 
external data make improvements in Z (statoscopic and solar methods) and in Y (solar 
method). However the three experimented methods have not improved the distribution of 
the errors in X nor have lessened the entity in any apparent way. 
4. A very important research project of the Commission deals with the comparative 
study of the adjustment methods: this study has been ony partially executed. 
At first the Commission decided to apply the same adjustment procedure to all the 
bridgings, so that the results could be compared. The main aim of the adjustment was, 
therefore, to reduce the errors in order to allow, in an easy way, a comparison amongst 
the residual errors. Afterwards, we thought it useful to accept the offer made by two 
Centers, namely: the Institut Géographique Militaire of Bruxelles and the Bundesamt