Additional notes. 
Since the y-parallax method for the determination of radial distortion 
in aerial photographs according to Hallert has been treated in this paper 
and the measurements have been performed by Dr. B. Hallert, the 
editor has given him the opportunity to comment on the results. 
SOME COMMENTS ON P. THAM: VERTICAL PARALLAX- 
ES, LENS DISTORTION AND STEREOSCOPICAL MODEL 
By B. Hallert, Stockholm 70 
[irst it must again be emphasized that no parallax measurements (x 
or y) can be used for calibrations of survey cameras. Certain non-pro- 
jective systematic disturbances, primarily radial distortion, of the pho- 
tographs (the reconstructed bundles of rays) can however, conveniently 
be determined under normal circumstances, as proved in many practical 
tests, from parallaxes only. y-parallax measurements do not in prin- 
ciple require any control information on the ground and are excellent 
for routine checks of normal models. — Some very important limiting 
factors for the application of parallax methods in general must however 
always be noted. This was strongly emphasized in some papers to the 
Stockholm Congress, where the y-parallax method was published: "The 
Grid Method and the y-parallax Method for the Determination of Syst- 
ematic Disturbances in Aerial and Terrestrial Photographs", page 14— 
15, and "International Comparative Tests. A suggestion”, page 2, See 
Int. Archives of Photogrammetry Vol. XII 1956. A detailed treatment 
of the limitations was made in: "Results of y-parallax Measurements in 
a Wild Stereocomparator and Some Additional Notes to the y-parallax 
Method for the Determination of Some Types of Systematic Distur- 
bances in Aerial Photographs". Photogrammetria 1958—59, Nr 2, 
section 3. Only the grid method is completely reliable and should be 
used for all basic tests of cameras from the air. 
In the actual case the filter evidently has ruined the interior orienta- 
tion of the camera and has introduced correlations as stated in the re- 
servations. These conditions are abnormal. 
According to the last sentence in Dr. Tham's paper, all empirical 
relative orientation is more or less incorrect. This is doubtless true. 
This serious drawback of the usual photogrammetric procedure can 
evidently only be theoretically correctly eliminated by the numerical 
relative orientation method or by application of the principles of nu- 
merical corrections. The measured y-parallaxes are simply corrected 
with the distortion y-parallaxes p, in the orientation points before 
further computations.