remained subdued during the middle night hours and had 
little, if any, effect upon the midnight spectrum curve. By 
0400 hours the Click Chorus had increased in intensity 
again and probably was the controlling agent for the base 
noise spectrum at this hour. However, it now was some- 
what cyclic in nature and was less intense than at 2000 
hours. 
The question as to what the shallow hydrophone base 
spectrum around midnight [fig. 13(S)] should be attributed 
is unanswered at present. The Motorboating sound is 
probably responsible for the raised level in the 150-to-500 c/s 
region, since it was present in many of the samples 
representing this hour and in a wide range of intensities. 
But the cause of the raised spectrum below 150 c/s remains 
a mystery. 
In figure 13(D), the dashed-line base represents the 
minimum levels just before and after 2300 hours at the 
deep hydrophone. The curve for the peak in the base 
ambient noise occurring at approximately 2300 hours is 
compared here with the spectrum curve for the Rhythmic 
Grunt sound, which around this hour was a dominant drone 
at these frequencies. The Grunt whose spectrum is shown 
is from the shallow hydrophone output. Closer correspon- 
dence could have been obtained using a different Grunt 
sample, particularly one from the output of the deep hydro- 
phone. A typical grunt density during this period was 60 
to 80 distinct grunts per minute, which, assuming a typical 
grunt duration of 500 msec, leaves very little time un- 
occupied by pulses of this sound. Presumably indistinct 
Rhythmic Grunts were also present, filling in the ''silent"’ 
intervals. 
The shallow hydrophone picked up fewer grunts, but 
a high proportion of them had a good S/N. Some shallow 
hydrophone Rhythmic Grunts had the spectrum shown, but 
others had a spectrum very similar to the 2300 curve 
between 60 and 300 c/s. On the other hand, all the Rhythmic 
Grunts recorded from the deep hydrophone seemed to be of 
the type having good level at 100 c/s like the 2300 spectrum. 
