TIDAL DATUM PLANES 39 
In general, during periods when no change is made in the adjustment of the gage, 
the differences between staff and scale will be approximately constant. Any difference 
which stands out strikingly from the others should be rejected from the computation 
of the mean difference. 
It is to be observed that the differences between staff and scale will vary system- 
atically if the inlet to the float well becomes clogged. In that case the difference will 
be greater than the average for the rising tide and less than the average for the falling 
tide. These differences thus furnish a check on the proper functioning of the float well. 
With the determination of the corrected setting for the scale, the tabulation of the 
hourly heights of the tide and of the high and low waters is carried on as outlined in the 
preceding paragraphs for the tabulation of the record on cross-section paper. 
In connection with the tabulation of the tide record it is assumed that throughout 
the period of observations the tabulations are referred to a staff the zero of which is 
maintained at a fixed level. If during the period of observations the staff is changed, 
the height relation between. the two positions of the staff must be accurately deter- 
mined. Whenever possible it is preferable to take account of this change in staff in 
connection with the tabulation, so that the whole series may be referred to the same 
staff. However, it frequently happens that the tabulations must be made prior to the — 
determination of the exact relationship between the two staffs. In that case full ex- 
planation should be noted in the column of remarks of the hourly ordinates and high 
and low water tabulations; and, as soon as the correction necessary to reduce these 
readings to the zero of the previous staff is determined, this correction should be noted 
on the tabulated sheets. 
Irregularities in Tide Curves 
To secure a correct representation of the rise and fall of the tide, the inlet to the 
float well of an automatic tide gage is made sufficiently large to insure free communi- 
cation with the water outside the float well. If the inlet is too small, the tide curve 
will show a smaller range of tide and a retardation in the times of high and low water. 
As a result of making it large enough to insure free communication, disturbed conditions 
of the sea will be reflected by irregularities in the tide curve. An example of such 
irregularities is shown in Figure 18, which is a representation on a reduced scale of the 
San Francisco tide curve for November 21, 1910. 
Two kinds of irregularities in the tide curve are seen in Figure 18. The first con- 
sists of numerous small ‘saw teeth,’’ which appear throughout the curve and which 
show up particularly well between the 11th and 12th hours of the day. These “‘saw 
teeth” represent the rise and fall of the larger waves and ocean swells which enter San 
Francisco Bay from the ocean. The second irregularity consists of larger and slower 
fluctuations, which appear suddenly about 4:45 a. m. and continue with diminishing 
amplitude until about 10 o’clock. These fluctuations are brought about by stationary- 
wave oscillations of the water within the bay and are known as seiches. 
Seiches are brought about by various agencies. Heavy winds, sudden variations 
in barometric pressure, and seismic waves due to seaquakes—all these bring about 
seiche movements of the water which cause irregularities in the tide curve. The seiches 
_ shown in Figure 18 were caused by a rapid fall and rise of atmospheric pressure at San 
Francisco. In Figure 19 are shown seiches due to heavy winds. The curves of that 
figure reproduce the tide curves at Atlantic City, N. J., for the first 3 days of January 
