232 
PACIFIC SCIENCE, VoL XI, April, 1957 
eter, are at 418, 491, 552, and 630 m/z (see 
inserts in Fig. 2). 
Irradiance (H) was obtained from 
where n = recorder reading, E = energy of 
the source through the wave-length interval 
d X, T f = measured transmission of the indi- 
vidual interference filter through the wave- 
length interval d X, and S = the relative sensi- 
tivity of the photomultiplier tube through 
the wave-length interval d X. The value of k 
was obtained as indicated above. 
Tracings of the recordings are shown in 
Figure 2. Orange light (X max, 630 m/z) is 
seen to cut out at 11 meters, and no flashes 
are apparent. With the 418 m/z filter, the dark- 
current level of the instrument was reached at 
a depth of 280 meters. 
The 418 m/j, cast reveals an actual increase 
in irradiance with depth between 100 and 220 
meters. Sudden flashes again occurred be- 
tween 100 and 280 meters, but these, as re- 
corded, were of lower amplitude than those 
in the unfiltered cast, presumably because of 
the narrowness and reduced intensity of the 
transmitted band. 
These observations indicate that the light 
generated at depth is blue-violet (see Harvey, 
1940, for spectral distribution of biolumin- 
escence). It can be seen too (Fig. 2) that the 
flashes produced in the surface layers are 
greener. 
The greater depth attained with a 418 m/z 
filter in the instrument (cf. Figs. 1 and 2) and 
the deeper flashes may be explained on the 
assumption that some of the light-generating 
constituents of the population had by then 
sought greater depths as the moon cleared the 
mountains and approached its zenith. The 
possibility that these constituents comprise a 
photoregulated community similar to the 
Californian deep-scattering layer may be in- 
ferred from this assumed downward migra- 
tion with increasing illumination. 
This inference is substantiated by observa- 
tions made off San Diego (Lat. 32° 38' N., 
Long. 117° 35' W.) on January 17, 1956. At 
this time EDO fathograms and light measure- 
ments were made simultaneously during the 
periods of the twilight migration of the sonic- 
scattering layer. Figure 3A (1640-1650 hours) 
shows that flashes were first recorded as the 
photometer reached the depth of the scatter- 
ing layer. Thirty minutes later (1720-1728 
hours) the top of the scattering layer had 
reached a depth of 120 meters. Figure 3B 
shows that the first flashes recorded by the 
photometer were now at this same depth. On 
the day of these observations sunset was at 
1706 hours. 
We have recorded flashes, during our scat- 
tering-layer studies, in the Pacific, Atlantic 
and Mediterranean. They appear to increase 
both in frequency and amplitude during the 
twilight migration of the layer and at night. 
On this account it is extremely difficult to 
obtain an accurate picture of the spectral dis- 
tribution of moonlight in the sea, or even of 
sunlight at scattering-layer depths. 
Since the completion of this paper we have 
heard from Professor George L. Clarke of 
Harvard University (Clarke and Backus, in 
preparation) that he too has recorded flashes 
of luminescence off the Florida coast. 
This phase of work on the sonic-scattering 
layers received additional support from the 
National Science Foundation and the Pauley 
Fund of the University of California. The in- 
quiry was pursued from the Marine Labora- 
tory of the University of Hawaii. The Califor- 
nian data were obtained from the "USS 
EPCE(R) 855.” Acknowledgment is made of 
the generous loan of the Research Vessel 
"Makua” by Mr. Vernon Brock, Director of 
the Territorial Fish and Game Commission, 
Hawaii. The U. S. Air Force and Marine 
Corps were unstinting in their loan of equip- 
ment and facilities. Data were analyzed and 
most of this paper was written at The Labora- 
tory, Plymouth, England. The hospitality of 
the director, Mr. F. S. Russell, F. R. S., is 
most gratefully acknowledged. Professors 
