Notes on the shrinkage of photographic emulsions 
on film and glass plates 
By H. Kasper and J. M. Zarzycki (Heerbrugg) 
The constant development of new methods for strip radial triangulation, and the 
economic need for maintaining precision while surveying as large an area as possible, 
soon ran into an obstacle constituted by the inadequacies of the pictures. 
This shortcoming was removed by the new WILD lenses. The standard f/4-2 Aviotar 
and the wide angle f/5:6 Aviogon, which are used with film and plates in the new RC5 
and RC7 aerial cameras respectively, far surpass the types of lenses known until 1948, 
not only in resolving power, but also as regards the other practical requirements of the 
photogrammetrist. 
But the precision of the plotting instruments also increased markedly during the past 
four years. 
Due to this, various sources of error, which formerly were only of secondary impor- 
tance, were traced with greater accuracy and certainty, and the need for their 
elimination became very obvious. One of these is the distortion of the emulsion layer. 
For these reasons, the much discussed question of film shrinkage must once again be 
taken up in detail. 
The following is a report on experiments which were conducted (in part) in cooperation 
with the Eidgenössische Landestopographie (Swiss Fed. Top. Service). 
The experiments were designed to investigate the properties (as to shrinkage) of the 
emulsions used on: 
a) Gevaert-Dia-Plates 19 x 19 cm (7:48 x 7-48") 
b) Kodak Super XX-Film (Aerographic safety base) 
c) Ilford Aerofilm 
) Gevaert-Avi-Panchro Microgran 27? Film (Aerographic base) 
e 
) Gevaert-Avi-Panchro-Hyperrapid 33? Film (Aerographic base) 
) Agfa Aeropanfilm 
In order to check on the dimensional stability of the emulsion on glass plates, a 
e 
+ 
precision grid 19 x 19 cm (7-48") was copied on ten Gevaert diapositive plates, of which 
an additional copy (one each) was made on diapositive plates. 
The grid was checked with a precision measuring instrument, to + 1-5 microns. It 
provided a standard which was almost free from errors, but the corrections it showed 
were further verified by computation anyway. 
The twenty control plates were measured like those carrying the grid, with a mean 
accuracy of + 1:5 microns. 
In order to make the same checks on the films, the test grid was copied once again on 
film strips 19 cm wide (1-48") ten times each, at a distance of 1 m (40^). 
The film grid was then copied on glass, in order to get an exact record of it for given 
temperature and humidity conditions, and so as to avoid any distortion of the film 
during measurements. These diapositive grid plates were measured in the same manner 
as the original grid plates.