1915] Michael, et al. : Hiidrographic Records of Scripps Institution 19 



Assuming the sextant to be properly adjusted, the following sources 

 of error are important : 



1. Difficulty in identifying the points of bearing with those on the 

 charts. This difficulty arises from variation in their appearance due 

 to different positions of the observer. 



2. Difficulty in clearly distinguishing the point.s becaiLse of the 

 hazy condition of the atmosphere. 



3. Impossibility of measuring both angles simultaneously, which 

 introduces error because of the drifting of the boat. 



Experience gained through repeated iLse of the same three points 

 of bearing may reduce the first source of error sufficiently to make its 

 effect negligible. This is especially true of our work in the region 

 about Los Coronados Islands, where the error arising from the otlier 

 two sources is estimated by Captain Crandall to have varied between 

 0.05 and 0.3 miles, depending largely upon the condition of the atmos- 

 phere and the distance from the islands. 



The pelorus consists of a circular plate mounted horizontally in 

 gimballs and placed at some point on board ship affording a clear 

 view for taking bearings. Revolving about the center of the plate are : 

 (1) a dumb compass card, usually engraved on metal, whose face is 

 level with the raised periphery of the plate, on which score lines are 

 marked indicating true directions from its center parallel and per- 

 pendicular to the keel of the ship; (2) a pivoted horizontal bar car- 

 rying at its extremities a pair of sight vanes so arranged that the line 

 of sight always passes through the vertical axis of the pelorus, and 

 having an index showing the point at which the line of sight cuts the 

 dumb compass. 



In using the pelorus to determine the location of the boat a definite 

 course is held. The instrument is then set on a line parallel with the 

 ship 's keel and the sight vanes are arranged so as to give the compass- 

 bearing on first one point of land and then vipon a second and, as a 

 check, upon as many more points as may be desired. 



Aside from the error due to faulty identification of the points of 

 land with those on the charts, large errors are due to the following 

 causes : 



1. Difficulty of holding the boat on the same course, without making 

 too great headway, wliile the sight vanes are set successively upon 

 the points of bearing. 



2. DiflSculty of getting clear sights while the boat is rocking and 

 pitching. 



