(396) 
As we see the influence of instrument-scale is considerable. In order to get 
the best values as well for the error in terrain as also for the error in the image 
plan, we have to take the instrument-scale as large as possible. The greatest 
influence seems to be caused on both instruments by smallest irregularities of 
the coordinate systems and their readings, the influence is due to the instrument. 
It is obvious that it affects the result more, if the instrument scale is small. 
It is to add that the material used for the investigation was of medium 
quality concerning the photographs, they were taken on film and printed on 
plates. The plates were used for the measurements. 
The control points were white-washed boundary stones and pegs, fixing 
of them by ground-measurement and making them visible was not performed 
for the scope of this investigation of accuracy but they were fairly good also 
for this purpose. The Autograph Wild A 5 was not in state of best adjustment 
and on the stereoplanigraph the measuring was performed not with lenses of 
the Aviotar type, which was used for taking the photographs, but with lenses 
of the Geodar type. Between these two types a difference of distortion exists, 
their value is about 0,01 mm. All these influences affect the general accuracy 
to some extent, it is therefore not possible to get an exact comparison 
between the two instruments used. Such a comparison was not the aim of 
the investigation. But the influence of instrument scale was proved rather 
clearly, as the different values of m, reached with the different instrument 
scales can not be cleared up by other causes. It seems necessary to test the results 
attained by our investigation still more thoroughly in order to get a deeper 
insight into the instrument-errors and to evaluate the standard values m, under 
the different conditions of plotting. 
The value m, = 0,017 mm got from the measurements using an instrument- 
scale of 1:5000 shows the increase of accuracy reached within the last years. As 
this value includes all error-influences also these of fixing and identifying the 
control points and as the general conditions of measuring and plotting were not 
the best these good result seems to be of interest.