me 
  
inclination determining quantities surveyed on the ground. Thére is a considerable 
risk of systematic (regional) elevation errors. These regional errors in the Z-coor- 
dinates are often much more dangerous and decisive for the use of the map than 
a corresponding regional error in the XY -coordinates. 
From the premisses mentioned above one must start from the fact that in 
careful photogrammetric mapping the elevation points within the mapping district 
must be densened to such an extent that each stereo model can be levelled. 
As levelling is the cheapest part of surveying") the cost of these complemen- 
tary levellings usually amounts to only a small part of the total mapping costs. 
The purpose of the photogrammetric triangulation work is then reduced to 
the determination of planimetric points (XY -coordinates). Thanks to the fact that 
it is possible to level the stereo models, the accuracy of the determination of the 
XY -coordinates is considerably increased, specially when the geodetic control net 
Is sparse. 
II. The principle of the method. 
IL1. The essence of the method is that each stereo model included in the trian- 
gulation strip is separately oriented in a stereo instrument, at which the base is set 
in the same position (e.g. *base inwards"). The levelling of the models is done 
through geodetic quantities (e.g. elevation points) ?). 
The scale of the first stereo model is determined by the aid of a known 
distance etc. The other models receive a preliminary scale at the plotting in the 
instrument by successive transmission of scale (graphical or numerical) between the 
models. In each stereo model coordinates for control points and connection points 
are determined in the model system. The latter are planimetrically well definable 
points, situated in the areas in common with the adjoining models. The coordi- 
nation of the different model systems to one system (the O-system) and the fitting 
in of the latter into the geodetic system is numerically executed by transformations. 
To obtain an accurate result from this method of triangulation and to make 
it generally fit for use two primary problems must be solved. 
IL.2. Tbe choice of connection points. 
To transfer the scale between the models through *connection-bases" does not 
give the same accuracy theoretically as the usual transference by restoring the 
heights of common points. The greatest disadvantage of this method is that it is 
difficult to find sufficiently definable connection points in terrain lacking sharply 
definable details (forests, certain field areas, archipelagoes etc.). 
This problem was solved by marking two or more connection points as thin 
crosses in the central line of every negative. The measuring mark of the stereo- 
instrument is stereoscopically adjusted on the surface of the model where the 
cross is placed on one of the pictures, whereafter it is monocularly adjusted on the 
1) It is possible to lower the costs still more by using the lately constructed levelling-instruments 
with so called selfhorizontalizing line of sighting. For average and small scales the elevation 
point determination would be considerably facilitated by using the levelling carriage constructed 
in the U.S.A., which is towed by a car or a jeep. 
?) By using the Finnish horizon method at the photography to determine the differences in 
tilt, or by using a spirit level at the plotting to restore the tilt of the common picture, the 
demand for elevation points in each model would be reduced. 
2 
  
    
   
  
  
  
  
  
  
  
  
  
   
    
    
  
  
  
  
  
  
  
  
  
   
    
  
  
  
  
  
  
   
   
   
  
  
  
    
    
crc 
as 
is 
scc 
th 
ar