An initial experiment (Fig. 4) was carried out using two "Tesa" 
displacement gauges, one on each face of a kinematically supported 
photographic plate in a humidity controlled enclosure. Summing the two 
displacements gives the average displacement of the plate, and differencing 
them gives the change in thickness of the plate. When the humidity was 
changed, the "Tesa" gauges showed the direction of the bowing, and its 
extent, but not its form. The constant difference between the two 
displacements also showed that the effect was a true bowing of the plate, and 
not a change in thickness of the emulsion, causing an apparent bow. Several 
different types of plate of varying thickness were tried, together with a 
plain sheet of glass to act as a control. The results are summarised in the 
table below: 
Type of plate Thickness Bowing for 75% Relative 
Humidity change 
Aviphot 1.50mm 35um over 100mm 
1.55mm 19um over 100mm 
Agfa 10ET75 1. 30mm 18um over 100mm 
2.25mm 8um over 100mm 
Plain glass 1.50mm less than 0.1um over 100mm 
Table 2 
These results showed that there was a serious effect with normal 
photographic plates. The form of the distortion was measured by holography 
(see Fig. 5). Firstly a hologram was made of the rear surface of a 
photographic plate, which had been spray-painted matt white to provide a 
diffusely reflecting surface (the rear surface was chosen as the paint would 
have sealed the emulsion surface to moisture). The hologram was then replaced 
and the test plate viewed by looking through the hologram, with the laser 
illuminating both the hologram and the plate in exactly the same way as when 
the hologram was taken. Interference fringes could then be seen between the 
real test plate and the reconstructed image from the hologram. With the test 
plate in the same humidity conditions as when the hologram was made, no 
fringes were visible, but when the humidity was changed fringes appeared 
indicating the form and extent of the bowing. Plates 2 & 3 show results from 
a typical Agfa 10E75 plate showing the degree of bowing (0.33um per fringe) 
and its form (very nearly uniform curvature). Again a control experiment was 
performed using a blank sheet of glass, which showed no change of shape with 
humidity, showing that the effect is due to the emulsion. 
From the table above it can be seen that the bowing is worse for thinner 
plates, and worse for Aviphot plates than for 10E75 plates. This is as would 
be expected since Aviphot is a much thicker emulsion than 10E75. The use of 
thick plates (say 6mm) with a thin emulsion, such as 10E75, would greatly 
reduce the humidity bowing Such a thick plate would also be much less prone 
to deformation due to clamping forces in the plate holder. An even better 
solution to the plate bowing problem would be to use a photosensitive 
material not sensitive to moisture, such as photoresist or 
photothermoplastic. 
192