lively discussion. Some people are of the opinion that the altitude should 
be reduced to 450 metres in order to achieve sufficient accuracy in the 
ma ps . 
Experimental work 
In order to discover partly what accuracy could be achieved as regards 
horizontal control points if these were acquired by photogrammetric 
600 
metres or 450 metres — and the choice of film — colour film or pan- 
  
triangulation, partly what effect the choice of flying altitude 
chromatic film — would have on the result of the plotting, a series of 
different tests was carried out. 
Empirical determination of the accuarcy of model triangulation 
Test object: An arbitrarily chosen area of about 230 hectares (36 map 
sheets in the scale 1: 400). 
This area had been photographed from an altitude of 600 metres 
(scale of negative 1: 4000) with panchromatic estar film in three strips 
with five models in each strip. Within the area there were about seventy 
signalled points with known coordinates. 
Measurements 
The stereo models were measured in the stereoautograph Wild A7, 
and coordinates were measured for all the signalled points and 
a sufficient number of natural connection points. The coordinates were 
registered with a Wild Ek5a. 
The stereoautograph was checked both by the measuring of grids 
according to Hallert-Kaasila and by the measuring of grid models. The 
first mentioned measuring gave the standard errors of unit weight 
3.5 um and 3.4 um for the projector concerned and insignificant 
Y-Breiten and X-Schiefe errors. The grid model measuring resulted in 
the standard errors s, — 2.3 um, s, — 3.9 um and ss, = 32 jm. 
(All values in the negative scale). 
Calculations 
The block transformations were carried out according to AB Geocode's 
programme P 412. Various calculations were carried out. 
l. Eleven well distributed control points. No corrections for systematic 
Crrors. 
213