4415 HYDROGRAPHIC MANUAL PaGE 338 



disappears. If it does not, try reversing the magnet in its tube. Careful steering is required, as the 

 effect of yawing may be mistaken for that due to heeling error. The heeling error is due in part to' 

 the fact that the iron beams of the vessel are no longer horizontal when the vessel is heeled over. 



(10) The compensation for semicircular and quadrantal deviation made with magnets and spheres 

 will not hold good for all latitudes. If the vessel is operated in a considerable range of latitude, a 

 Flinders bar will be required to compensate for changes due to change in latitude. The Flinders bar 

 is usually located forward of the binnacle. Until opportunity is afforded to determine the exact 

 length required, it is advisable to insert one of sufficient length to reduce the deviation on east and 

 west headings to 10° before compensation, i. e., with the fore-and-aft magnets removed. A Flinders 

 bar, so selected, will not entirely counteract the changes in the deviations due to change in latitude 

 but will minimize them considerably. 



4415. Determination of Deviation 



The principal methods of determining the deviation of a ship compass for various 

 headings of the ship are: 



(a) Bearings of the sun. 



(6) Bearings of a range whose azimuth is known. 



(c) Bearings of a distant object. 



(d) Reciprocal beaftngs. 



These methods are described in detail in Special Publication No. 96 and in the 

 American Practical Navigator (Bowditch) and need not be repeated in this Manual. 



(e) Another method is a modified form of method (6) in which neither the azimuth of the range 

 nor the magnetic variation need be known and the two objects comprising the range need not even 

 be charted. The ship swing is made as usual with both right and left rudder, the compass bearing 

 of the range being measured as it is crossed on each heading. The two compass bearings of the range 

 for identical headings are averaged. The mean values for all headings are then averaged and the 

 difference between this average and the mean compass bearing for any heading will be the deviation 

 of the compass for that particular heading. 



A ship swing to determine the deviations of the compass shall be made at the 

 beginning of each season's work, after a long lay-up period, after undergoing repairs at 

 a shipyard, after a change of working ground involving a considerable change in 

 latitude, and during the season if the deviations are found to differ appreciably from 

 those on the deviation card. (See 144.) 



When swinging ship, the vessel must be on an even keel and the sea comparatively 

 calm. To obtain the best results, good steering is necessary, and the ship must be 

 steadied on each heading 2 to 3 minutes before the observations are made. The varia- 

 tion should be accurately known. It is printed on the compass rose of the chart for a 

 particular locality, but should be corrected to the date of the swing by applying the 

 annual change. 



The deviation can be determined most accurately by swingmg ship around the 

 end of a wooden pier, if there are no magnetic metals in the vicinity. By using bow 

 and stern spring lines the vessel can be warped around the sides and end of the pier 

 and held stationary for the observations on each selected heading through 180°. Then 

 the vessel must be turned around and the maneuvers repeated for the headings through 

 the other 180°. The observations can be made on the sun or a distant object whose 

 azimuth from the end of the pier is known or can be determined. Because the water is 

 calm and the vessel can be held absolutely steady on each heading, an excellent swing 

 should result. 



A ship swing should not be made during a magnetic storm. The occurrence of 

 these is unpredictable; tliey may last several days, are associated with sun spots and 



