The new Type VI binnacle, with fore-and-aft and athwartships magnets, makes compass correction by the tentative methods of Special Circular No. 55 a comparatively simple matter. Nevertheless, the navigator, in his first compass adjustment aboard a new ship, will find one or two small practical details missing, which will consume time in ascertaining, but which, if once tabulated, will enable him to keep the compass approximately correct all the time, even without observations.
Suppose, for instance, with head east, he finds by an azimuth that there is a westerly deviation of 4°. What does he do? He moves the fore-and-aft magnets a small amount up or down and notices the effect. After two or three trials, he gets them at the right height. Suppose he had a table of constants in which he found that for the given latitude, raising the magnets one-tenth of an inch made 2° easterly deviation on east. He would at once, under the supposed case, raise the magnets two-tenths of an inch.
With a similar tabulated value of the constant for the athwartship magnets, head north, he would at once correct the observed deviation on north. Similarly, suppose it has been found that moving out the quadrantal correctors one inch makes about 22 easterly deviation on NE. That gives him the means of correcting the quadrantal deviation. The compass is now approximately correct. A swing for residuals will either show no appreciable deviation, or, should there be any, will enable him at once, with the aid of the constants already employed, to make such small additional changes as will leave the compass correct on all courses.
In practice, after the compass has once been approximately corrected, the simplest way to re-adjust is to swing on eight points, finding the deviation on each. Then, to eliminate A, take the mean on north and south (changing the sign on south), which determines how much to change the athwartship magnets; and the mean on east and west (changing the sign on west), which determines how much to change the fore-and-aft magnets. To determine how much to change the quadrantal correctors, take the mean on NE., SE., SW., and NW. (changing the sign on SE. and NW.). The compass should now be practically correct, but it will be well to swing on the same points for residuals, as a final check.
Much time will be saved in this work by a simple modification of a detail of the tentative method. Instead of computing magnetic azimuths for given stated times in advance, it is quite as easy to make a curve on cross-section paper in which the ordinates are minutes of local apparent time, and the abscissae degrees of magnetic azimuth, that is, true azimuths corrected for variation. Then for any given instant of time, the magnetic azimuth may be picked off directly from the curve. The difference between the magnetic azimuth and the compass azimuth is the deviation. The first swing should take 32 minutes (4 x 8). With 10 minutes to change the correctors and 32 minutes for the second swing, the whole operation should be finished in an hour and a quarter. A time-azimuth curve, as actually used aboard the Massachusetts, is shown in Fig. 1.
Next, it will be found convenient to keep tabulated a complete record of all changes in the corrections. The columns for this table are as follows: Date, place, latitude, height of fore-and-aft magnets, height of athwartship magnets, distance of quadrantal correctors, height of heeling magnet. In the course of a year or so of cruising, it will be found that a latitude curve may be sketched in for each corrector; that is, the average position of the corrector will be shown for each latitude. Such curves, taken from actual experience aboard the Massachusetts, are shown in Figs. 2 and 3.
Theoretically, the quadrantal correctors are supposed to be fixed in position after one careful adjustment. In practice this is not found to be the case. Aboard the Massachusetts they have to be changed from 11 ¼ inches to 12 ¼ inches in changing from 17° to 43°, north latitude. It is possible that the correctors are not actually pure soft iron, but have acquired some small amount Of magnetism. For this reason they should be marked starboard and port and kept always on their own side and turned the same side to the compass.
The Massachusetts is supplied with the old type starboard-angle binnacles, which necessitates a slight modification in the practice as above described. As it happens, the starboard-angle is approximately 180°. Hence raising or lowering the magnet tray affects the deviation on east and west, but not on north and south. The magnets are actually fore-and-aft magnets, and hence may be treated as described above. Instead of raising and lowering athwartship magnets to effect changes on north and south, the method aboard the Massachusetts is to rotate the tray very slightly, thus changing the starboard-angle. The constant is the amount of deviation caused on north by a change of one degree in the starboard-angle.
The heeling magnet may be placed by any of the usual methods. A good method, as used aboard the Massachusetts, was as follows: With the ship heading north, she was listed several degrees to starboard by training the after turret, observing continuous azimuths before and after. She was then brought on an even keel and afterwards listed in the same way to port, continuing the azimuths. This was repeated with the heeling magnet in different positions, until the continuous azimuths when plotted to form a time azimuth curve showed as a smooth curve. With deviations due to a heeling error, the time-azimuth curve would show as an irregular figure lying on either side of the smooth curve, and would have to be smoothed out by changing the height of the magnet.
An example will now be given, taken from actual work, to illustrate the method. On December 28, 1901, off Nipe Bay, Cuba, in Lat. 210 10' N., the Massachusetts was swung on the cardinal points as follows: (See Fig. 1.)
N. (+) 0.7° Ely.
E. (+) 2.5° Ely.
S. (—) 0.3° Wly.
W. (—) 1.4° Wly.
No attempt was made to correct the small deviations found on north and south. The mean on east and west (sign changed on west) was (+) 2.0° Ely. From the table of constants it is seen that raising the magnet tray one tooth of the ratchet makes about 2° Ely. deviation on east; hence it would have to be lowered one tooth to correct 2° Ely. deviation. The tray was accordingly lowered from a total of 18 teeth to 19 teeth. On the next swing these results were obtained:
N. (—) 0.2° Wly.
E. (+) 0.7° Ely.
S. (+) 0.2° Ely.
W. (+) 0.2° Ely.
Next the ship was put on the intercardinal points, except NE., which could not be obtained on account of the shadow of the mast, with these results:
SE. (—) 0.7° Wly.
SW. (+) 1.3° Ely.
NW. (—) 1.8° Wly.
The mean of SE. and NW. gave about (—) 1.2° Wly. To correct. The constant for moving the spheres is about 2 ½° deviation on NE. for 1 inch increase in the distance. On SE. the constant would be 2 ½° Wly. deviation. Hence to correct 1.20 Wly. deviation, the spheres would have to be moved in towards the center one-half inch. This was done with the following results on the next swing:
SE. (+) 0.6° Ely.
SW. (—) 0.1° Wly.
NW. (—) 0.3° Wly.
The compass was now practically correct on all courses. In this case the semi-circular deviation was corrected before attempting the quadrantal. Ordinarily it will do equally well to swing on eight points and correct both together. The second swing will then give the residuals. The time required for this compensation was one hour and fifteen minutes.
After the latitude curves are once determined, the compass may be kept nearly correct all the time. This is useful in case of fog, or when making land after a considerable change of latitude when there has been no opportunity to swing. Azimuths must be taken as frequently as possible whether the compass is correct or not. No change from the usual practice is intended in that regard. Careful records aboard the Massachusetts indicate that the ship's permanent magnetism does not suffer any material change. There are slight temporary changes after target practice, and after lying for some months at the navy yard on the same heading. This is the so-called retained magnetism, but it disappears after a few days of ordinary cruising. In changing the correctors according to the latitude curves, it is usually advisable to keep a day behind, on account of the retained magnetism, the ship tending to retain for a short time the induced magnetism of the latitude left.
As a final example, the Massachusetts made a trip from Pensacola, Lat. 30° N., to Hampton Roads, Lat. 37° N., in April, 1903. After arrival at the latter place the magnet tray was set at a height of 15 teeth of the ratchet below its highest position, and the quadrantal correctors were set at 11 3/4 inches from the center. These positions are in accordance with the curves of Figs. 2 and 3. On April 28, three days after arrival, the ship was swung on eight points in latitude 36° 52' N., with the following results:
N. (+) 0.1 0 Ely.
NE. (—) 04° Wly.
E. (—) 0.7° Wly.
SE. (—) 0.30 Wly.
S. (—) 0.2° Wly.
SW. (—) 0.1° Wly.
W. (+) 1.1° Ely.
NW. (+) 0.7° Ely.
The mean quadrantal was (— 0.4° + 0.3° — 0.1° — 0.7°)/4 = 0.22°, which could have been corrected, had it been desired, by moving the spheres out something less than one-tenth of an inch. In practice it has been found convenient not to change the spheres less than an eighth at a time.
The mean on north and south did not require correction. On east and west the mean was (--0.7° - 1.1°)/2 = --0.9°. This could have been corrected by raising the tray one-half a tooth of the ratchet. In other words, the adjustment was as close as the appliances would permit. However, a small shim piece has been provided, which may be secured to the toe of the pawl, making the equivalent of one-half a tooth of the ratchet. By this means the tray was raised from 15 teeth to 14 ½ teeth, leaving the compass correct on every point within a deviation of less than one degree.
These curves are now habitually used aboard the Massachusetts with almost uniform results. Azimuths are always taken daily on whatever course steered, but principally as a check only. When the ship is long enough in one place, or time affords, a swing is made on eight points for residuals; and, by means of the table of constants, such small changes are made as may be shown to be necessary.