whereafter a second laser flash can be generated (curve d 
and el. 
The foregoing applies only for long interval times (100 to 
700 us). If this time becomes too short to store the energy 
for the second pulse, we must close the shutter again when 
half the stored energy is used, leaving the other half 
available. This situation is represented in curve f and g 
and. is typical for interval times between 1 and 100 ys. 
If the output of the energy monitor Mon 1 is coupled to.a 
storage oscilloscope with adjustable trigger delay via inte- 
grating electronics, one can see a trace as shown in Fig.3 
in double pulse operation. The vertical parts correspond to 
a calibrated energy, the time interval to the horizontal part 
between both pulses. 
The action of the amplifiers is similar to that of the 
oscillator. The flash tube delay times however are chosen so 
that the maximal energy becomes available at the moment be- 
tween both pulses. In practice, the output is monitored via 
Mon 2 and the delays set for maximum. The energy of the 
flash pulses can be adjusted by changing the voltage over the 
amplifier flash tubes. The output lens L2 is interchangeable 
and must be adapted to the area to be illuminated. 
3) SET-UP. 
The use of the holocamera was abandonned in most experiments, 
as it was originally designed for illuminating structures of 
around 2 x 2 m, requiring large distances to the holographic 
plate. We needed much shorter distances as our subjects were 
rather small (head, hand, arm and chest). Besides, a large 
set-up flexibility is wanted for this field of applications. 
The easiest set-up is shown on Fig.4. Only one diverging beam 
is used, H is the hologram plate, S is a screen to avoid back 
illumination of H, and M is a mirror. By deplacing M laterally 
the correct beam ratio of 1/1 can be obtained. The whole 
set-up was made on an ordinary table. 
The uniformity of the intensity of our apparatus leaves much 
to be desired, probably due to thermal effects, impurities 
and/ar parasitic reflections. Spatial filtering is very compli- 
cated, as the energy density in the focal point of lenses 
easily surpasses breakdown of air, and was not attempted. 
Thanks to extensive work done at NPL (ref.4) we were able to 
produce a suitable scatter plate (D, Fig.4) giving minimal 
attenuation of intensity and change in divergence. 
To avoid backreflections in the laser, the scatter plate must 
be slightly inclined and to ensure sufficiently fine speckle 
on the subject, it must be mounted some distance away from the 
output lens. These parameters must be determined first in a 
suitable set-up with a He Ne laser (a and e, Fig.5). The diver- 
gence angle o of the pulsed laser must also be known. A glass 
master diffusor with equivalent divergence angle was produced 
by smoking glass with the vapour of ammonium chloride. 
The final scatter plate was made by exposing a holographic 
plate H (Agfa Gevaert 8E75HD), with a He-Ne laser beam of 
102