fe of the Coast of the United States on the Gulf of Mexico. 5 
taken for a comparison of the lunitidal intervals to trace the 
tee of the diurnal wave. The variations from day to = 
i ing less than the probable irregularities in the times of hig 
¥ water and the uncertainties in the observations; these means 
; give suitable numbers for comparison. The result would not 
i, have been greatly different had only a few of the observations at 
either end of the declination period been thrown off, but after 
examination we found these numbers to present apparently the, 
best results. 
At four of the stations, namely, Key West, Fort Morgan, Cat 
Island, and Galveston, hourly observations were continued during 
a year and upwards, and the decompositions in all the cases but 
Cat Island embrace that period. The annual change of diurnal 
establishment is very clearly seen in all these cases and is shown 
in the diagram No. 8. The law of the change is beget de- 
veloped in the larger tides of the Western Coast and, as deduced 
from the San Francisco observations, is shown upon the same 
diagram. In all the cases the actual computed results for the 
different halfmonthly periods are represented by the broken 
lines on the diagram and the line representing the curve is drawn 
with a free hand among the points. The general resemblance of 
these curves with however different maximum ordinates, is very 
striking, showing that the law of the change is the same, only 
the coéfficients of the fractions varying. f 
n the diagram of the San Francisco results the curve derived 
So as to use the mean of the two periods of six months. 
At the other fourteen stations on the Gulf of Mexico the ob- 
servations were continued from one to three lunations, and fell 
in different parts of the year. To reduce these erefure to the 
| = same period of the year it is necessary to emplo: the data from 
the localities where the whole annual change was embraced. The 
| results are plotted on the several diagrams, those from the Brazos 
to South West Pass on the curve from Galveston, those from the 
South West Pass to St. George’s on the curve from Fort iy a 
e m 
the observed changes and in those deduced from, the other com- © 
parisons, at least there are no ter contradictions than those 
ese 
the result to the mean of the year are deduced and the annexed 
table shows the diurnal interval as deduced directly from the ob- 
wee 
= 
