E
154 RELATIVE ORIENTATION IN MOUNTAINOUS TERRAIN, DISCUSSION
mately flat terrain, and the formulae for this
type of terrain which was present in the Monti
di Revôira experiment.
As you can see from the table there are
three elements which are unchanged: the dby.,
the dz» and dg». The corrections came out for
rough as well as for flat terrain, depending
upon the actual shape of the terrain in that
case.
Then we have the other two elements: the
db. and do». There are differences, but for the
standard error of unit weight soy there were no
differences which could be noted. Consequent-
ly, in that case we used the formulae for flat
terrain for the computation of certain of the
elements of orientation and also for the deter-
mination of the standard error of unit weight
which defines the accuracy of the relative
orientation after an adjustment.
Further we had the opportunity to test the
formula for error-propagation from the relative
orientation through the absolute orientation to
the final results of the plotting. As you can see
from Table 1 in the Additional Remarks to
the General Report of the sub-Commission
IV:4, there is very good agreement between
the theoretical and the practical results, al-
though in that case we used nothing but the
mean square value of the measured y-paral-
laxes in nine points as an expression for the
sy here. This means an approximation, yet the
results agreed so well, but I certainly hesitated
to publish them from the beginning.
With these words I just wanted to say that
the actual shape of the terrain is very impor-
tant, of course. Monti di Revôira represents
just one of the infinite number of possible
shapes of the ground.
Mr A. J. VAN DER WEELE: I should like to
thank Mr Hallert for his contribution, particu-
larly with regard to the formulae for the com-
parison of flat and mountainous terrain.
Is there anyone from the floor who wishes
to make additional remarks or who has a
question?
As this does not seem to be the case, I
declare this session closed as far as this par-
ticular subject is concerned.