66 University of California Publications in Zoology [ Vou. 14 
and bottom water. It will be noted that the mean annual salinity 
ranges from less than 16 per mille in Carquinez Strait, which re- 
ceives the discharge of the rivers, to over 31 per mille, just within 
the Golden Gate. The rise of these curves from Carquinez Strait to 
the latter point is fairly regular, save for a marked fall in both which 
is encountered in passing from San Pablo Bay into the main body of 
San Francisco Bay, i.e., from station H-4981 to H-4974. This might 
easily be interpreted as due to the admixture at this point of con- 
siderable quantities of fresh water, but no adequate source of fresh 
water is known here and the phenomenon may readily be explained 
otherwise. 
As already stated (pp. 11, 12), the upper, middle and lower regions 
of the bay were worked on different days. Now station H—4974 lies 
at the upper end of the middle series of stations, while H-4981 lies 
at the lower end of the upper series. Consequently, on the flood-tide 
work observations for the middle region were commenced at the 
former station, whereas the observations for the upper region were 
ended at the latter. It is needless to state that water taken at the 
commencement of the flood period will have a lower salinity for any 
given point than water taken toward its close. Conversely, at the 
time of the ebb-tide work the observations for the upper region were 
commenced at H-4981, while the observations of the middle region 
were ended at H-4974. Water taken at the end of the ebb period will 
naturally have a lower salinity than water taken at the beginning of 
the ebb period. 
Thus in both eases the conditions under which the work was per- 
formed led to the taking of water samples of a lower salinity at station 
H-4974 than at station H-4981. The peculiarities of our curves are 
thus adequately accounted for. Moreover, the same considerations 
show us how cautious we must be in comparing the mean salinity of 
any two stations unless we know that the two were visited at approxi- 
mately the same stage of the tide. This same source of error will be 
be discussed under another form later (pp. 71, 72). 
From the Golden Gate, the salinity curve naturally falls toward 
the lower end of the bay. Commencing with the first station below 
San Francisco, there is, however, no significant decline shown by the 
surface water, though such a decline is to be noted for some distance 
in the ease of the bottom water. The slight rise of both of these curves 
at the extreme southern end is probably of significance, since it is 
manifested by four of the six seasonal curves in figure M. We have 
