(474) 
The following types of film were tested at the laboratories: 
Kodak Super-XX 
Gevaert MICROGRAN 
Gevaert HYPERRAPID 
Agfacolor 
Grids were printed on the films under parallel light. The films were then 
developed and air-dried under normal conditions. Subsequently, the films were 
kept in the open in ordinary offices for certain periods of time so that they 
were exposed to normal atmospheric fluctuations. At certain time intervals, 
contact diapositives on glass plates were made of these grids, also under parallel 
light, and their coordinates measured with the Zeiss Universal Measuring 
Microscope. The accuracy of the measurements was + 1.6 u. The 63 grid points 
had already been numbered on the original so that the comparison with the 
grid printed on the film could be referred to identical points and irregularities 
of the original could be eliminated. The close network of the reference points 
also permitted to ascertain the interior behaviour of shrinkage. 
The measured coordinates of the film grids were transposed arithmetically 
by Helmert transformation to the corresponding points of the original grid 
and the results of the deviations represented in the form of “point errors” in a 
vector diagram. 
The results of the laboratory tests are as follows: 
1. The regular film shrinkage which is characterized by the scale factor in the 
transformation formulae, fluctuates according to time and the individual 
test film between +1.16%0 and —1.7°/00. 
2. The irregular film shrinkage which adversely affects the plotting accuracy 
consists essentially of the affine shrinkage lengthwise and widthwise to the 
direction of manufacture of the film as well as accidental local irregulari- 
ties. The average difference between lengthwise and widthwise shrinkage 
ranges for all films between 0.05?/o) and 0.18°/00. 
For the orientation of stereo pairs the affine shrinkage factor is non- 
existant owing to the elimination of the y parallaxes. Summarizing the data 
given by film by the same manufacturer, we find: 
" Mean" Error Maximum Error 
Kodak Super-XX 5H 124 
Gevaert MICROGRAN +5 9 
Gevaert HYPERRAPID + 4 u 184 
Agfacolor + 4u 18 u 
This goes to show that there is no essential difference between different 
types of film stock. 
Slightly improved and especially more homogeneous results were obtained 
by quick drying of the films with hot air blowers. 
In order to ascertain whether large film rolls have different shrinkage pro- 
perties, grids were printed on a 52 m film strip at intervals of 9, 17 and 26 m. 
The table shows that the results are similar to those of short pieces. Conser- 
vation in rolls or single photographs therefore seems to give the same results. 
A final clarification of this problem can be obtained only after the material has 
been tested for a period of several years.