120 
CHRISTIE. 
which was not determined by levelings to bench-marks; 
and the longer series was the more defective in this respect. 
There may be local peculiarities affecting the mean range 
of this tide on an earth not absolutely rigid, but it is incon¬ 
ceivable that the period should differ from station to station. 
Comparing the longitudes, dates, and epochs, a period of 
d. d. 
432.6 ± 3.2 
would make the waves identical, a very gratifying result, 
and one to which some weight might be given ; but it suffices 
to take the simple mean of the periods determined from the 
independent series, which is much the stronger value, and 
write 
d d 
Period = 431 zb 4 
where the probable error is doubtless a very fair measure of 
the precision of the determination. If we neglect any possi¬ 
ble local influence upon the range, and a here insensible 
difference due to difference of latitude, we may take the 
mean and write for both stations the half range as 
ft. in. mm. 
a v = 0.049 = 0.59 = 15 = O'. 144 
± 6 ±7 ±2 ±18 
and this, considering the character of the observations, seems 
a just conclusion from the data. Hence, neglecting an im¬ 
material adjustment of the epochs, we may write for the first 
term of the latitude-variation tide, as deduced from tidal 
observations at these two stations, the following values: 
SAN FRANCISCO. 
mm. 
151 _ ( 360° , ( 123°\ 1 
± 2 } cos { 431^4 t ~{± 12 ) } 
t reckoned in days from September 21, 1872. 
Minimum, 1872, July 15 ± 15. 
Maximum, 1873, February 15 ± 15. 
