il 
| 
| 
| 
| 
| 
  
  
  
Figure 1 
taken to obtain a symmetrical loading (Fig. 3 and 4). The heights in the measuring points 
were determined by the aid of the Leitz Ultropac incident microscope (Fig. 4). 
For the glass plate in the original plateholder (Fig. 1 C) a strong bending under 
loaded condition was found (Fig. 5). This was caused by the fact that the springs in the 
plateholders (Fig. 1,3) were so short that their reaction to the load from the “camera 
  
  
  
  
  
  
  
  
  
  
] 
28 *58 +88 | 
:27 57 87 | | 
Load : Load 
+26 *56 *86 | | 
. . . . . LI LI . " " | 
ji 45 25 fs dé 6s Camera body ; 
*24 +54 *84 | Photographic plate 
TUE = a 
:23 *53 83 | CR 7 
| Spring- 
+22 +52 2 . | | 
) | 
Figure 2. Controll glass plate in plate- Figure 3. Cross section through 
holder, scale 1:4. Control points are filled and loaded original plate- 
intersections in a 20 mm grid. holder, scale 1 : 2. 
body” (Fig. 1 A) caused quite a large bending moment in the glass plate. Due to that the 
springs mentioned were replaced by longer ones, giving remarkably better results (Fig. 
5, 4 and 5). As they still did not appear to be long enough, they were once more replaced 
(Fig. 1, B 2). 
The final plateholder springs were made so long, given such a shape and placed so 
in the plateholders that under loaded condition (during exposure) there should be no 
more risk of any disturbing bending moments in the negative plate. 
As a single measure for the flatness of the glass plate for the various test conditions, 
the mean square value of height deviations from best estimated horizontal plane are given