- 3 - 
.n use, so 
L in this 
i was about 
-up as 
) photograph 
•cury lamp 
tee between 
>per surface 
;wice the 
i phase 
’uctive 
1 is an 
)er will 
} air space 
. Thus a 
it level of 
)1 between 
Fig.3 shows 
Ls that it 
lumber of 
does not 
>e is 
mt, is not 
is already 
xrovides a 
)ut does not 
is the 
it changes 
Since 
ntion 
i’ringes 
50 at points 
serpretation 
A, B, C and 
i trough, 
, Leaving 
mal, leading 
¡.e-point, and 
$en by the 
msiderations 
'es of the 
;, A or B, 
The use 
in this case 
’ space 
;he inter- 
Ln interest 
of the investigation lay in finding the magnitude of the departures from 
flatness irrespective of sign, monochromatic light was used in most of the 
records. 
A count of the fringes between A and B in Fig. 3 will show that the height 
difference is about 60 half-wavelengths, i.e. 16 microns. On some inter- 
ferograms places have been found with as much as one hundred and ten fringes 
height difference (i.e. 30 microns) between points within the dimensions of one 
reseau square. It is clear therefore that there is scope for improvement in 
the arrangements for film-flattening in the register glass type of camera. 
The illustration in Fig. 3 is of the kind of pattern that is established 
some considerable time after the film has been moved into position and the 
pressure applied. It was noticed at an early stage in the experiments however, 
that the fringe pattern had not become completely steady even several minutes 
after applying the pressure. It was therefore decided to make a cine record 
of the course of events during a complete cycle of operation of the camera. 
A 16 mm film* was made and reveals that no interference pattern whatever is 
visible, even in monochromatic light, for a second or so after pressure is 
applied to the film. During this interval, therefore, the air space is 
appreciably greater than the figures already mentioned, although the degree of 
flatness of the film is not known; it is probably quite poor. After about 
three seconds from completion of film transport the general nature of the inter- 
ferogram is established, but considerable changes in detail continue for several 
seconds longer; smaller changes are still occurring after many minutes. 
These experiments point to the presence of air trapped under the film as being 
responsible for the major departures from flatness. 
It was decided, in view of these results, to investigate the effectiveness 
of the suction method of holding the film flat. In this case it is necessary, 
in order to use the interferogram technique, to arrange for an optically flat 
glass surface to be suitably supported at such a distance from the emulsion as 
to give a good fringe pattern without at the same time disturbing the film in 
any way. This is not easy, but the E.IoD. Laboratories at Harefield came to 
our assistance and succeeded in obtaining suitable interferograms• Meanwhile, 
an exploration, over small areas, of variations in the level of the film surface 
was made at R 0 A 0 E 0 with a depth measuring microscope. The results of the two 
investigations agreed satisfactorily, the interferograms showing maximum 
deviations of about 18 microns and the microscope 17 microns. 
There are one or two points of interest about the results of the suction 
methods of flattening. The first is that the interference patterns obtained 
are remarkably consistent from one exposure to the next, whereas with the 
pressure pad no such consistency is observed. The second is that the flattening 
is established within one second. It was noted also that, using a plate with 
many small holes, but no communicating channels, the areas of film between the 
holes were bowed in such a way as to suggest the presence of trapped air. 
!, "‘To be shown at the Congress