44 University of California Publications in Zoology [ Vou. 14 
perhaps also partially explainable by the influence of the overlying 
air. The latter, during the summer months, commonly has a consid- 
erably higher temperature than that of the water, while in winter this 
condition is reversed, the water being warmer than the air. 
Feb. Mor. Apr. May Jun. Jul Aug. Sept. Oct. Nov. Dec. Jan. 
Fig. E. Annual temperature curves for entire bay. The continuous line 
is based on flood figures, the broken line on ebb ones. 
Figure E, like the preceding one, represents the annual temper- 
ature cycle, but in the present case the two curves show the conditions 
during the flood and ebb tides respectively. It will be noted that the 
flood figures, for all but one period, indicate higher temperatures than 
do the ebb figures, and that in the single exceptional case the differ- 
ence is insignificant. For the year as a whole, the mean of the flood 
figures is nearly 0°5 C higher than for the ebb figures (Table 9). 
Moreover, this relation is found to hold for each of the stations with- 
out exception, and to hold for the surface water as well as for the 
bottom water. 
On first thought it would seem paradoxical that the water should 
have a higher temperature during the flood tide than during the ebb. 
The former water might be supposed to come more directly from the 
ocean than the latter, and the ocean, throughout most of the year, has 
been shown to have a lower temperature than the bay. The ease is 
exactly paraliel to that which will be discussed in the chapter on salin- 
ity (pp. 70-72), and the cause in the two cases is probably the same. 
The comparison which our figures afford is not exactly a comparison 
between the flood and the ebb tide as a whole, but a comparison be- 
tween the early flood and the early ebb. 
As was stated above, the experiment was twice made of anchoring 
the ship in the course of the main current which passes through the 
