NAVIGATION NAUTICAL ASTRONOMY. 



[COMPASS ERRORS. 



namely, 155 Iff 57', instead of 34 43 3', as we 

 havo Uken it above ; and we too that by tubtratting 

 96 68' we get the tune result namely, 58 47' 59*. 



Whatever method of solving the problem of double 



altitudes be employed, there U always, in particular 



instance*, a degree of uncertainty as to whether the sum 



:iference of a particular pair of arcs or angles U to be 



taken, and which uncertainty can be removed only by 



uoe to the latitude by account. 



The method proposed by Mr. Ivory, and which has 

 been put in a very commodious form by Mr. Kiddle, in 

 his Navigation and Nautical Astronomy, is perhaps the 

 shortest of the correct processes, when the object ob- 

 served is the sun ; but the investigation of it is very 

 long and complicated, and the several steps of the work 

 are far leas easy of recollection than those above : it has 

 the advantage, however, of deciding the ambiguity here 

 mentioned in the case of the sun, with but very little 

 trouble. 



But in the case of a pair of stars, the uncertainty is 

 not so easily removed : an amount of calculation, equiv- 

 alent to a repetition of the work involved in the last step 

 above, has to be gone through. 



Now it occurs to us, that the best and most satisfactory 

 way of coming at the proper value of the angle P S Z, is 

 to directly compute this angle (in imitation of step 2 

 before) from the three sides of the triangle P S X 

 namely the co-declination P S, the co-altitude Z S, and 

 the co-latitude by account P Z. This extra work may 

 be regarded as at least an approximate verification of the 

 entire process up to the end of step 3 ; and, in executing 

 it, seconds in the several arcs need not be attended to : 

 it will be sufficient to take each side to the nearest 

 minute. 



The resulting value of the angle P S Z would be a safe 

 guide to the value of it, which step 3 ought to give ; and 

 the concluding step might then be worked without any 

 misgiving. 



It may be further remarked here, in reference to a 

 double altitude of the tun, that when the true co-latitude 

 P Z is obtained, we may combine it with co-declination 

 P S, and the co-altitude Z S, to determine the hour- 

 angle Z P S, the apparent time from noon when the alti- 

 tude furthest from the meridian was taken : the cor- 

 rection for the equation of time being applied, the result 

 will be the mean time at the ship when S was observed. 

 The chronometer, proper allowance being made for its 

 daily loss or gain, will show the mean time at Green- 

 wich ; and the difference between the two will be the 

 longitude in time. 



I '.ut this important subject will be more fully discussed 

 in the next chapter ; we shall only add here, that an 

 altitude near the meridian is not eligible for determining 

 tintf, because, in that position, altitudes vary so slowly, 

 that a small error in altitude may occasion a compara- 

 tively largo error in time. 



E sample 2. In latitude 58 N. by account, the true 

 altitude of Capella was 69 23' ; and at the same time the 

 tnii- altitude of Sirius was 10 19' ; the following parti- 

 culars were also given by the Nautical Almanac : 

 Right Ascension of Capella 5h. 2m. 18s. Dec., 46 47' N. 

 Sirius 6h. 36m. 33s. 1627'S. 



Required the latitude. Ans. Latitude, 57 & N. 



THE DEVIATION OF THE COMPASS. The mariner's 

 compass is subject to two kinds of disturbance, which 

 operate against its true north and south direction. One 

 U a local disturbance attributable to the attractive in- 

 iliu-iico of the iron in the ship, which, when unequally dis- 

 tributed in reference to the magnetic meridian, or the 

 Uoe along which the compass would otherwise settle, 

 causes a deviation from that lino. The effect of this local 

 attraction is called the deviation of the compass. The 

 amount of the angular disturbance or aberration is ascer- 

 tained by experiment, before the ship proceeds on her 

 voyage ; and a table U then formed, which gives the 

 proper correction for different positions of the ship's 

 bead. But Mr. Barlow, of the Royal Military Academy, 

 Woolwich, ban devised an apparatus for counteracting the 

 effect of local attraction : a small iron plate U so placed 



near the compass, vertically and behind it, as that its 

 influence upon it may neutralise the action of the iron 

 in the ship; the proper situation for the plate being 

 ascertained by experiment when the ship in in li irlxnir. 

 For an account of this contrivance, and the mode of 

 applying it, the reader is referred to Barlow's Essay on 

 Magnetic Attraction!. It may be observed, however, 

 that much care is requisite in fixing the plate ; and that 

 if the distribution of the iron in the ship bo afterwards 

 changed, the plate mnst be readjusted ; but this incon- 

 venience must attach to every method of estimating t ho 

 deviation from the results of observations previously 

 made, under particular conditions, as to the amount and 

 position of the disturbing materials. Local deviation has 

 sometimes been found to be so great, that, in a ship loaded 

 with iron, the course by compass has been known to 

 differ from the proper course by so much as 30.* 



THE VARIATION or THE COMPASS. The vertical plane 

 passing through the direction of the horizontal needle at 

 any place, marks out the magnetic meridian of that place, 

 just as the vertical plane through the north and south 

 points of the horizon marks out the geographical meri 1 i.-ui 

 of the place. The angle traced on the horizontal plane 

 by the vertical planes just mentioned, measures the 

 variation of the compass at the place. There are but very 

 few places on the surface of the earth where the needle 

 points directly to the north and south points of the 

 horizon, or where the variation is zero. The line in 

 which such places lie is called the line of no variation, 

 and is a curve of very complicated form. It was not till 

 probably some centuries after the discovery of the com- 

 pass that the needle was suspected ever to lie out of the 

 plane of the geographical meridian : it seems first to have 

 been observed by Columbus, in 1492. 



The variation of the compass is not only different at 

 different places ; but, what is more remarkable, it is not 

 constant even at the same place. At London the vari- 

 ation was eastward till about the middle of the seven- 

 teenth century ; in 1659 the needle pointed due north 

 and south, the variation being ZITO, so that at that time I 

 the magnetic coincided with the geographical meridian. 

 After this the variation became westerly, and it continued 

 to increase till the year 1818, when it appears to have 

 attained it greatest limit of westerly variation namely, 

 24 3<y. Since then the variation has been slowly 

 diminishing, and it is now about 23 vxst. 



In order to ascertain the variation of the compass at 

 any place, it is necessary that we find by computation, 

 or, by some means independent of the compass, the true 

 bearing of a celestial object ; then observe the magnetic 

 or compass bearing ; the difference of the two will be the 

 variation of the compass ; including, however, the local 

 deviation, unless this have been neutralised, as explained 

 above. 



There are two forms of the problem : the object whose 

 bearing is observed may be either in the horizon or above 

 it. In the former case it is the magnetic amplitude that 

 is observed ; in the latter case it is the azimuth : in 

 both cases the latitude of the place is supposed to be 

 known. 



To FIND THE VARIATION FROM AN OBSERVED AMPLI- 

 TUDE. It has already been seen that the angle which 

 the equinoctial makes with the horizon of any place is 

 the co-latitude of that place (page 1074). This angle is 

 therefore given ; and as the object observed is in the 

 horizon, and its declination or distance from the equi- 

 noctial known, wo have given the perpendicular (tin- 

 declination) and the opposite angle (the co-altitude) of a 

 right-angled spherical triangle to find the hypotenuse 

 (the tnii- amplitude); and the difference between this 

 and the observed amplitude is the variation, including 

 the local deviation when not previously counteracted. 

 Refraction causes objects to appear in the horizon when, 

 on an average, they are 33' below it ; consequently, the 

 compass-amplitude should be taken when the sun's 

 centre, or the star selected, is about :;:;' /.'< tlm ilij>, 

 above the sea-horizon ; or, allowing 16' for the sun's 



ThU rabjcct, unil that of the Variation, U fully entered Into In the 

 KcUon on MagnetUm in this volume. ED. 



