15 
sx = s 0 } / Q> 
etc. 
(46) 
Consequently the standard errors of the corrected coordinates can be 
found for arbitrary numbers and locations of the stars. Then the standard 
error of the angles can be computed from the principal distance of the photo 
graph and the positions of the actual points. It should be noted that in this 
simplified procedure regular errors may remain undetected to influence the 
results. The determination and comparison of the standard errors of unit 
weight with the corresponding values from calibration procedures will, how 
ever, indicate certain regular errors. Normal distribution tests of the residuals 
are also advisable. The adjustment procedure used here is similar to that 
known as a Helmert-transformation. The determination of the geometrical 
quality, which is the main purpose of this paper, has been treated by the 
present author in previous papers in photogrammetry, see for instance Hallert 
1960 a. 
2. Reconstruction of rays with the aid of the interior orientation 
of the photograph 
In this case the rays are reconstructed from measured image coordinates and 
the elements of the interior orientation only. The procedure is illustrated in Fig. 2. 
Fig. 2. The direction toward P is to be determined with the aid of the elements of the interior 
orientation only. These elements are indicated as the position of the principal point H' 
in the negative plane and the principal distance (camera constant) c. Attention must 
also be paid to possible regular errors of the image coordinates as determined through 
the camera calibration. A statistical value of the irregular errors should also be estimated.