288 
Fig. 13. Comparison of chlorinity for a year at sta- 
tions 1, 5, and 10. Chlorinity of surface waters shown 
by broken line, of bottom waters by solid line. 
The surface layers were especially affected 
by these rainstorms. In the shallow stations, 
5, 6, 7, 10, and 11, where only one sample 
was taken, chlorinities dropped to about the 
same degree as the surface waters of the deep- 
er stations such as 1 and 2. The surface layer 
of brackish water lost its separate identity 
after about a month, and deep and shallow 
water became almost homogeneous but of a 
lowered chlorinity. In the winter season of 
1949-50, about 2 months were required for 
the chlorinity values to resume a normal level 
after the heavy rains. 
In the 24-hour cycles, no diurnal changes in 
chlorinity were observed which could be cor- 
related with either day and night or with tides 
(Fig. 14). 
Oxygen samples were determined by the 
Winkler method for dissolved oxygen (Hol- 
lister, 1950: see section, "Determination of 
Dissolved Oxygen in Fresh and Sea Water”; 
Kolthofif and Sandell, op. cit., pp. 614-619)- 
All oxygen values were graphed to show sea- 
sonal variations and to compare some of the 
stations (Fig. 15). Averages were used when 
the differential between top and bottom was 
PACIFIC SCIENCE, Vol. VII, July, 1953 
Fig. 14. Comparison of chlorinity for three diurnal 
observations. Chlorinity of surface waters shown by 
broken line, of bottom waters by solid line. Part of 
August samples were lost, hence break in graph. 
less than 0.25 milliliters of oxygen at normal 
temperature and pressure per liter of sea water 
at 20°C. From the beginning of sampling to 
mid-July, 1949, sampling was done in the 
early afternoon with the few exceptions noted. 
These afternoon values were, on the whole, 
higher than the morning values. Values ranged 
from 1.10 milliliters per liter to 7.27 milliliters 
per liter, these extreme values being the ex- 
ception rather than the rule. On the whole, 
no marked seasonal pattern was detected, and 
diurnal changes were more noticeable. 
Oxygen concentration displayed a definite 
diurnal change (Fig. 16) in all three 24-hour 
cycles, reaching a minimum in the morning 
between 0500 and 0700 and a maximum in 
the afternoon between 1400 and 1800. This 
diurnal change is presumably related to the 
biochemical activity of plants producing an 
excess of oxygen in hours of daylight and the 
consumption of oxygen by both plants and 
animals in the hours of darkness. 
Saturation values were determined (com- 
putations based on Table 24, Harvey, 1928: 
