Microscopte Structures of Wind Waves 
If we designate r as the radius of the focal spot of the teles- 
cope on the water surface, the degree of saturation of the light signal 
(s = 1 for saturated signal) can be shown as 
di 
2125 Joos"! = -= oer]? (4) 
where 
hla sina 
is the half-width of the bright portion of the water surface image (see 
figure 3a). Of course, the radius of the focal spot of the telescope on 
the water surface, r, depends upon the diameter of the pinhole, d, as 
well as on the characteristics of the telescope lens. The calibrated 
response for the present setup, as well as those obtained directly 
from (3) and (4), is plotted in figure 3b. The scattering of the calibra- 
tion points is believed to be the result of local deflects on the calibra- 
tion cylinder. 
In summary, the light signal is continuous and saturated to 
a prescribed level, insensitive to the change of curvature, as long 
as the angular change of the wavy water surface from the downwind 
face to the upwind face is less than the acceptance width of the instru- 
ment (about 1°), As the angular change increases beyond that, or if 
the surface curvature increases further, the signal becomes discon- 
tinuous. The signal is essentially a light pulse. The intensity of the 
signal, i.e., the pulse height, is related to the surface curvature; 
the period of the signal, or the pulse width, is the time required for 
a detectable slope to pass completely under the instrument. 
The distribution of the surface slopes is determined by accu- 
mulating the numbers of light pulses for each inclination of the instru- 
ment for a constant period (10 min). To determine the distribution of 
surface curvatures, each series of light pulses is sorted according 
to their intensity into 50 channels of the pulse-height analyzer, with 
preset intensity bands. The output from the analyzer, the height dis- 
tribution of light pulses for a given instrument inclination, is first 
traced on a x-y plotter and later digitized. 
The block diagram of the apparatus is shown in figure 4a. 
The light source is a 1500 W incandescent lamp, approximately 18 cm 
long. The photomultiplier tube is a nine stage, side-on unit with 
S-4 spectral response, The high voltage power supply is adjustable 
so that adequate sensitivity with minimum dark current noise can be 
obtained. The output of the electrometer is comprised of irregularly 
1439 
