64 U. S. COAST AND GEODETIC SURVEY 
Atlantic, Gulf and Pacific coasts of the United States furnish illustrative examples of 
the necessity for specifying the epoch used. 
For New York Harbor there are available 56 years of observations, from 1893 
through 1948. This permits three 19-year series, 1893-1911, 1912-1930 and 1930-1948, 
the last two having the year 1930 in common. For the series 1912-30, sea level referred 
to a number of bench marks in the vicinity of the tide station was 0.09 foot higher than 
for the series of 1893-1911; for 1930-1948 it was 0.29 foot higher than for 1893-1911, 
and 0.20 foot higher than for 1912-1930. 
At Baltimore the tide observations cover the 46 year period 1903-1948. Taking 
the first and last 19-year series, 1903-1921 and 1930-1948, we find the latter 0.26 foot 
higher. For comparison with New York, if we form two like 19 year series, 1912-1930 
and 1930-1948, we find that sea level from the latter series was 0.23 foot higher as against 
a difference of 0.20 foot for New York. ane 
For Galveston the 40 years of observations from 1909 permit two independent 19 
year series, 1909-1927 and 1930-1948. Sea level derived from the latter series is 
0.39 foot higher than from the former series. Figure 31 indicates an accelerated rise 
of sea level at Galveston since 1940, which if maintained, will bring about even greater 
differences for subsequent 19 year series. 
For the three Pacific coast tide stations it will be sufficient to take comparable 19 year 
series at each of the stations, namely, 1906-1924 and 1930-1948. At San Diego the | 
latter series gives sea level 0.14 foot higher; at San Francisco, 0.16 foot higher; at 
Seattle, 0.12 foot higher. 
Specifying the epoch on which a given primary determination of mean sea level 
is based, permits correlation with mean sea level determinations made at other times, 
provided adequate bench marks are maintained. . 
Secondary Determination 
Observations covering a period of 19 years for primary determinations of mean sea 
level are required at but few places on the coast. At all other places a satisfactory 
secondary determination of this datum plane can be made by means of observations 
covering much shorter periods if the results are corrected to a mean value by com- 
parison with the primary determination at some suitably located tide station. The 
precision with which mean sea level can be derived by a secondary determination from 
various periods of tide observations can best be illustrated by examples. 
Day.—Since weather conditions at widely separated places may be markedly 
different on the same day, it is obvious that, in deriving mean sea level at any point 
from one day of tide observations, comparison must be made with a near-by primary 
station at which the changes in sea level will be similar. 
Figure 23 shows that at New York during the month of October 1947 sea level was 
lowest on the 2d and highest on the 31st, the values of sea level on the tide staff being 
respectively 4.94 and 7.28 feet, a difference of 2.34 feet. Suppose that from the 
observations for each of these two days it is desired to derive a value of mean sea 
level for New York, using Atlantic City, about 100 miles away, as primary tide station 
for comparison, 
At Atlantic City for epoch 1930-1948, mean sea level on the tide staff reads 6.49 
feet, while for October 2 and 31 the daily sea levels read, respectively, 6.09 and 8.41 
