230 
PACIFIC SCIENCE, Vol. XI, April, 1957 
the instrument, and at times apparently 
single flashes reached a level of 2 X 10 -3 /i 
watts/cm 2 . 
A later cast, with the photometer still 
pointed downward, revealed that at sunrise 
(0652 hrs.) the flashes were first recorded at a 
depth of about 110 meters. 3 All facts known 
to us indicate that these flashes are generated 
in situ and we interpret them as due to lumin- 
escent organisms, possibly both nekton and 
plankton. 
Records of bioluminescence at depth were 
first made in Hawaii at midnight on the night 
of June 9, 1955. The moon, three days past 
the full, had just appeared over the moun- 
tains. Its angle of elevation at 2400 hours was 
19°. The instrument was lowered without 
color filters, and total irradiance, as seen by 
the instrument, was recorded. The sky was 
partly cloudy, but the moon remained clear 
throughout the observations. 
Figure 1 is a tracing of the actual recording 
made on the Leeds and Northrup Speedomax 
recorder. Total irradiance (H) at any depth was 
obtained from 
n = kH = / ESd\ 
o J 
where n — recorder reading, E = energy of 
the source through the wave-length interval 
d X, S — relative spectral sensitivity of the 
3 The possibility that these are artifacts due to in- 
strument failure is to be considered. It is improbable 
that distortion due to pressure affected the circuitry, 
for the phenomenon was not observed when the instru- 
ment had been lowered to greater depths during day- 
light. It is improbable, too, that the effect could have 
been caused by water leakage, for moisture in the sys- 
tem has been shown to raise the dark current to such 
a high level that the recorder is immediately sent off 
scale. Furthermore the fact that color filters placed in 
the light-path reduced the amplitude of recorded flashes 
discounts the possibility that loose electrical connec- 
tions were responsible. Another photometer, similar to 
the one employed here, has since shown that the flashes 
are not recorded when the sensitive element is covered 
but are when it is exposed immediately afterward at the 
same depth. Thus we believe that the instrument was 
sensing real light flashes. 
Flashes of this nature could not have been caused 
by tilt of the instrument since it was pointed downward. 
photomultiplier tube through the wave-length 
interval, d A (300-700 m/z). The value of k 
was obtained by calibration of the entire in- 
strument against a Bureau of Standards source. 
The inset shows the relative spectral sensi- 
tivity of the 931-A tube used. 
The dark-current level of the photometer 
was reached at 200 meters. From about 50 
meters downward numerous brilliant flashes 
of light were recorded. 
At some depths, for example 120-130 me- 
ters and 150-170 meters, the background 
light level appears to have remained constant 
or to have increased rather than to have de- 
creased with depth. 
Many types of luminescent marine animals, 
e.g., euphausiids, generate a steady glow of 
light for limited periods in contrast with the 
discrete flashes produced by others (Harvey, 
1940). Possibly the sustained light level (Fig. 
1) on which the flashes are superimposed at 
depths of 120-130 meters and 150-170 me- 
ters was due to a background glow produced 
by concentrations of such types, or by nu- 
merous flashes at varying distances. 
An alternative explanation of the flashes 
can be found by visualizing a distribution of 
steadily-glowing organisms similar to that of 
stars in the sky, and a flashlike increase in the 
level of light due simply to the instrument’s 
approach to and retreat from a glow during 
its descent. The rate of descent of the instru- 
ment was about 32 meters/minute. For this 
to be the true explanation, the glows would 
have had to be about 1 meter of depth apart, 
for the frequency of the flashes was about 
36/minute. 
Many luminescent organisms flash brightly 
when irritated mechanically or chemically. It 
is impossible at present to determine whether 
the observed flashes were due to spontaneous 
luminescence or whether the luminescence 
was intensified by agitation of the animals by 
the instrument. 
Four more casts were made, using narrow- 
band interference filters, during the forty-five 
minutes after midnight. The angle of eleva- 
