64 HANDBOOK OF MAGNETIC COMPASS 



the spheres are athwartship, the total E coefficient for the ship is 6° E., 

 as observed. Figure 31 indicates that the spheres in their present posi- 

 tion are correcting 6° E. D error, hence the total D coefficient of the 

 ship and Flinders bar is 10° E. Figure 33 indicates that 6° E. ^ and 

 10° E. /> coefficients require slewing the spheres 15i/^° clockwise from 

 their present athwartship position. The resultant quadrantal error 

 is indicated as ll.T°. Figure 31 indicates that the 7" spheres should 

 then be moved to the 11" position after slewing 151/2° clockwise so as 

 to correct both the D and E errors. Use of this chart will eliminate 

 mathematical or trial-and-error methods of adjustment for quadrantal 

 errors, as well as quickly provide information for physically moving 

 the spheres. 



95. Flinders bar adjustment. — As has previously been stated in 

 chapter VII, it is generally impossible to place the correct amount of 

 Flinders bar without reliable data obtained in two widely separated 

 magnetic latitudes. The placing of Flinders bar by the use of an 

 empirical amount, or by an inspection of the ship's structures, is 

 merely an approximation method and refinements will usually be 

 necessary when data is obtained. There are several methods of ac- 

 quiring and utilizing such latitude data in order to determine the 

 proper amount of Flinders bar, hence an elaboration on the following 

 items : 



(1) The data necessary for calculation of Flinders bar length, 

 and the conditions under which this data should be acquired. 



(2) The best method of utilizing such data to determine the 

 proper length of Flinders bar. 



96. Data required for Flinders bar adjustment. — The data re- 

 quired for correct Flinders bar adjustment consists of accurate tables 

 of deviations with details of corrector conditions at two different mag- 

 netic latitudes, the farther apart the better. See figure 30 for example 

 of how such data is recorded on NBS Form 1104. Should it be impos- 

 sible to swing ship for a complete table of deviations, the deviations 

 on east and west magnetic headings would be helpful. On many occa- 

 sions ship's log data is available, but is of little use for Flinders bar 

 calculation because it is not reliable. The following precautions should 

 be observed when such data is to be taken in order to assure that 

 observed deviation changes are due only to changes in the H and Z 

 components of the earth's field. 



(1) Degaussing should be secured, by a reversal process if nec- 

 essary, at both latitudes before data is taken. 



(2) If the ship has been docked or steaming on one heading for 

 several days prior to the taking of these data, the resulting tem- 

 porary magnetism (Gaussin error) would create erroneous devia- 



