Joty—Altitude of Celestial Object at Sea at Night-Time. 561 
The altitude so determined, of course, requires a larger sub- 
tractive correction for “‘dip” than would be required when using 
the visible horizon. 
+ S31 
Fig. 1. 
Let AH be the sensible horizon or tangent plane to the Harth’s 
surface at the observer’s position, and AJ’ the line touching the 
visible horizon vertically beneath the star S,, and reaching the eye 
of the observer, whose height of eye above sea-level is OA: also 
let AS be the line extending from the observer’s eye to the signal, 
which last is also vertically beneath the star. The usual correc- 
tion for dip is the angle HAH"; while that required by the use of 
the signal is the angle HAS. The angle HAS, of course, exceeds 
the normal dip by an amount depending on the proximity of the 
ship to the signal. Deducting it from the observed reading, 
S,A8 on the sextant (after this is corrected for instrumental 
error), the remainder S,4 ZH is the elevation of the object above the 
sensible horizon. The line OS follows, of course, the curvature of 
the Earth’s surface. A short table of dips for objects nearer than 
the visible horizon is given in Norie’s “ Navigation,” and in other 
works on navigation. A more extended table is desirable for the 
purpose of the present method. Such a table should read at least 
to minutes of arc. 
In smooth water results of considerable accuracy would pro- 
bably be obtained by the observation as described. If attention 
be paid to obtaining verticality of the star over the signal, and to 
obtaining a good reading of distance, and the ship’s way be reduced 
