JoLty—A liitude of Celestial Object at Sea at Night-Time. 565 
usual manner. The mariner may, however, require to find his 
position by a Sumner problem. If a second known star is avail- 
able, preferably not far from the first star, and he devotes a little 
longer time to the observation, this may be accomplished, the 
final accuracy depending chiefly on the state of the sea and the 
care with which he reads his distance and his angles. 
The nature of the procedure in taking the altitudes of two 
stars to work a Sumner is as follows :—Having, as just described, 
taken the angular elevation of a star, S,, above the signal, the ship 
is now put in such a course as will bring the second known star, S,, 
vertically over the signal, while at the same time preserving 
unaltered the distance between the ship and the signal. The 
angular elevation of the second star above the signal is then 
observed, and the same correction for dips applied, when all the 
data for the problem are obtained. 
+52 
Fig. 3. 
The figure, showing the successive positions of the ship pro- 
jected on the horizontal, will explain the principles on which the 
ship’s course is determined in going from the position where the 
first observation is made to that in which the second is made. 
Sis the signal ; S,, the first star ; A, the position of the ship when 
reading the angular elevation of S,, above S as viewed from the 
ship. ‘The second star, S., is seen in the direction 48, from the 
ship at A. 
The observation of altitude at A being made, the bearings of 
the two stars are observed. The angle a is thus obtained. Now, 
if B is the required second position of the ship, the star S, will be 
sighted vertically over the signal, and the line BS, will be parallel 
