[99] WORK OF FISH COMMISSION STEAMER ALBATROSS. 215 



rithm of tlie Leight, and from the sum subtract the logarithm of the 

 number of divisions. 



The accuracy of this short method is shown in the following example, 



Avhich is an extreme case : 



Example. 



A light-house 200 feet high is found to subtend an angle of 1,700 M. 

 D. The value of one division of the instrument in this vessel is 3".655, 

 of which the log. cotangent is 4.7515377, or in practice 4.75154. 



Sliort method. 



200 feet log. 2.30103 



1 M. D log. cot. 4.75154 



7. 05257 

 1, 700M. D.- log. 3.23045 



Eigorous method. 



3".655 X l,700 = 6,213".5 = lo 43' 33".5. 



200 feet log. 2.30103 



1° 43' 33".5 log. cot. 1. 52097 



6,637.4 feet log. 3.82200 



6,639.3 feet ..log. 3.82212 



The smaller the angle, of course the smaller will be the discrepancy. 



For rapid work in a hydrographic survey or reconnaissance 10 feet 

 is found to Ue a convenient leugth of staff to handle, and the logarithm 

 of 10 being 1.00000, makes the computation all the easier. A board 

 10 inches broad, painted white, with a 2-inch black stripe down the 

 middle, will be found to be an easily distinguished target. 



In the navigation of the Albatross and the location of dredging and 

 other stations, Sumner's method of finding the x>osition at sea is used 

 in extenso. All positions, however determined, are plotted as lines, 

 and not as points, the intersection of two such lines, corrected for the 

 intervening run and current, defining the exact position. The lines 

 of position consist of portions of circles of equal altitude of the sun, 

 moon, stars, and planets ; parallels of latitude deduced from meridian 

 or ex-meridian altitudes of the same bodies ; lines of bearings of head- 

 lands or well-known objects on shore ; circles of equal distance from 

 known objects, found by micrometer or by their dipping below the 

 horizon. 



Whenever practicable, the errors of the chronometer are found by 

 comparison with the time obtained by telegraphic connection with some 

 observatory clock. When such connection is not feasible, equal alti- 

 tudes of the sun are taken and computed by Chauvenet's method, 

 either for apparent noon or apparent midnight. 



The formula used for determining the latitude from the observed 

 meridian altitude of any heavenly body is — 



L=5;-fd 



For determining the latitude from an altitude near the meridian of 

 any body, the following formula has always been used : 



cos Sq = sin h + cos L cos d versin t, 

 in which the approximate value of L is used in comj^uting the term 



