230 



Papers from the Department of Marine Biology. 



it seems possible for clouds to make notches in the diurnal curves, and 

 sudden squalls have a similar effect. 



On July 10, 19, 20, 21, 22, 23, 24, and 25 very frequent determina- 

 tions were made, so that the notches in the curves could be studied. 

 With the omission of these small notches, the diurnal curves are plotted 

 in figure 6. It may be observed that there was a general upward drift 

 in temperature during July of about 2. As a consequence of rise in 

 temperature and increased cloudiness, there was a fall in the oxygen 



July 12345 



10 



FIG. 6. Diurnal variation curves of sea-water from station A, Tortugas, during July, 1917. 

 The date is marked on the right-hand border of the day interval, or where omitted it may 

 be determined by counting the waves of the curve. The trough of the wave represents the 

 early morning and the crest of the wave represents the early afternoon. The top curve 

 represents the number of cubic centimeters O* per liter; the second curve, the excess or 

 deficiency of saturation; the third curve, the number of cubic centimeters CO-z per liter; 

 the fourth curve, CO? tension; fifth, pH; and sixth, temperature. 



concentration in the sea-water during July. July 10 was an unusually 

 calm day (up to 9 p. m., when a brief squall appeared) and shows great 

 extremes in all determinations. Contrasted with this are the windy 

 days, July 13, 14, 15, and 16, during which the oxygen curves were very 

 much flattened out. This shows that the oxygen content of the sea- 

 water is somewhat influenced by the atmosphere. In figure 3 it may 

 be observed that the oxygen content of the water north of Hatteras 

 is increased with decreasing temperature and increasing absorption- 

 coefficient, which is another demonstration that the oxygen content 

 of the water is influenced by the air. On the other hand, shallow water 

 is not nearly in equilibrium with the air in regard to oxygen. At 



