Tam ULM PR MEE 
During very bad light and weather conditions — around freezing 
point, rain, snow and sometimes hard wind — about 130 signals, which 
could be regarded as remaining in their original positions from photo- 
graphy date, were surveyed with section and resection with theodolite 
by the staff from the collaborating offices. First the approximate coor- 
dinates were computed by the electronic computer and afterwards a 
rigorous adjustment was performed in the same computer. All the points 
were adjusted simultaneously according to a programme by Mr A Hä- 
kansson, M. S. Surveying. The point standard error was, after exclud- 
ing some points with large-errors, less than += 10 mm for the remaining 
100 points. In Fig. 4 the quality of the geodetic points — finally used 
— 1s indicated. 
The signals were levelled by the staff from the State Power Board 
with a Zeiss Ni II automatic levelling instrument. All the signals were 
levelled twice independently, and from the differences between the two 
Z-coordinates of each point the elevation standard error was calculated 
to + 7,5 mm. 
Photogrammetric measurements. Due to the time factor — it was 
assumed to be of interest that the preliminary results were to be report- 
ed to the London Congress — the number of orientations and models 
had to be limited as follows: 
The Hydrographic Department (= S) (operator A. Thunberg) trea- 
ted the models B and C in the Wild-Autograph A7 no. 310 at the 
Royal Inst. of Technology (instrument scale 1:667). 
The National Road Board (= V) (operator O. Grinndal) treated 
the model A in the Wild A8 no. 650 (instrument scale 1:800). 
The Geographical Survey Office (= K) (operator B. Persson) 
treated the models A, B and C in the A7 no. 515 with automatic data 
recording Wild EK3 (instrument scale 1:667) 
The diapositives were contact diapositives. 
The models were oriented relatively and absolutely in the normal way; 
the vertical parallaxes measured in 15 points and the x-, y- and z-coor- 
dinates recorded for 60—110 points in the various models A, B, and C. 
The pointing of the signals (a deformation caused by the movement of 
the aircraft could easily be seen) was made once but the procedure was 
also repeated in the opposite order for all the signals and the mean 
values were used. 
The mean square values of the residual y-parallaxes as well as their 
adjusted values in 15 points (according to Hallert's formulas) for the 
six oriented models called BS, CS, AK, BK, CK, AV are condensed in 
Table 1. 
The x- and y-coordinates of the signals of each model were trans- 
formed twice; on all the check-points (56—108) and on about 10 well 
distributed accurate points respectively with the aid of the electronic 
computer. As an example the results of the two treatments of model C 
  
4 
ee 4.