ADJUSTMENT PROCEDURE 51 



(2) If the above method is impractical the Flinders bar length 

 will have to be set approximately by : 



(a) Using an empirical amount of Flinders bar which has 



been found correct for other ships of similar structure. 



(&) Studying the arrangement of masts, stacks, and other 



vertical structures and estimating the Flinders-bar 



correction required. 



If none of these methods is suitable, the Flinders bar would best 

 be omitted until data are acquired. 



The iron sections of Flinders bar should be continuous and at the 

 top of the tube with the longest section at the top. Wooden spacers 

 are used at the bottom of the tube to achieve such spacing. 



78. Having adjusted the length of Flinders bar, place the spheres 

 on the bracket arms at the best approximate position. If the com- 

 pass has been adjusted previously, place the spheres at the best posi- 

 tion as indicated by the last deviation table. In the event the compass 

 has never been adjusted, place the spheres at midposition on the 

 bracket arms. 



79. The next adjustment is the positioning of the heeling magnet by 

 means of a properly balanced dip needle, as discussed in chapter XV. 



80. These three adjustments at dockside — Flinders bar, spheres, 

 and heeling magnet — will properly establish the conditions of mu- 

 tual induction and shielding on the compass, such that a minimum 

 of procedures at sea will complete the adjustment. 



Compass coil installations should be adjusted at dockside in order 

 to save time at sea. Chapter X discusses this procedure. 



81. Expected errors. — The "Summary of Compass Errors and 

 Adjustment," figure 19, lists six different coefficients or types of de- 

 viation errors with their causes and corresponding correctors. A 

 discussion of these coefficients follows: 



The A error is more generally caused by the miscalculation of azi- 

 muths or by physical misalignments, rather than magnetic effects of 

 unsymmetrical arrangements of horizontal soft iron. Thus if the 

 physical alignments are checked at dockside, and if care is exercised 

 in making all calculations, the A error will be rather insignificant. 

 On a standard compass where an azimuth or bearing circle is used to 

 determine deviations, any observed A error will be solely magnetic 

 A error. This results from the fact that such readings are taken 

 on the face of the compass card itself rather than at the lubber's line 

 of the compass. On a steering compass where deviations are obtained 

 by a comparison of the compass lubber's line reading with the ship's 

 magnetic head as determined by pelorus or gyro, any observed A error 

 may be a combination of magetnic A and mechanical A (misalign- 

 ment). These facts explain the procedure wherein only mechanical 



