• '» ^f-ii 
 
 849 
 62 
 
 
 
 
 IN AN IRON SHIP 
 
 J.WHITLY-DIXON 
 
BOUGHT WITH THE INtlOME 
 FROM THE 
 
 SAGE ENDOWMENT FUND 
 
 THE GIFT OF 
 
 iienrg W, Sage 
 
 1891 
 
 A:/i(^iSl INGINEERJNG LIBRAf^4i/9^a 
 
Cornell University Library 
 QC 849.D62 
 
 The mariner's compass In an iron ship; h 
 
 3 1924 004 599 365 
 
Cornell University 
 Library 
 
 The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924004599365 
 
THE MAEINER'S COMPASS 
 IN AN IRON SHIP. 
 
 HOW TO KEEP IT EFFICIENT, AND USE IT 
 INTELLIGENTLY. 
 
 WITH 
 
 SOME REMARKS ON ELECTRIC INSTALLATION IN ITS 
 RELATION TO THE COMPASS. 
 
 J. WHITLY DIXON, 
 
 Betired Captain, B.N. 
 
 WITH DIAGRAMS AND MAGNETIC CHAKTS. 
 
 PORTSMOUTH : 
 
 GEIFFIN & CO., 2, THE HAED, 
 
 PuWtsfjets bg Appointment to ?^er iHlajestg. 
 
 LONDON AGENTS: SIMPKIN, MARSHALL & CO. 
 
 1898. 
 [All rights reserved.'] 
 
 PRICE TWO SHILLINGS AND SIXPENCE. 
 
LONDON : 
 
 PRINTED B7 WILLIAM CLOWES AND SONS, LIMITED, 
 
 OTAUFOBD ffTREET AND CHABINO CROSH. 
 
PEEFACE. 
 
 This little work has been undertaken at the suggestion of 
 some of my brother officers now engaged in navigating her 
 Majesty's ships. Its aim is to give a general statement of 
 the facts and laws governing magnets and the earth's 
 magnetism ; to enable a seaman or reader to intelligently 
 comprehend a mariner's compass, and the reason and order 
 of its adjustment when disturbed by the iron of, or in, 
 a ship. 
 
 Many admirable books and pamphlets on magnetism and 
 the deviation of the compass exist, but the difficulty and 
 inconvenience of consulting them all are obvious. The 
 writer has endeavoured to compress in a small book all the 
 necessary information likely to be required by an officer 
 navigating one of her Majesty's ships or one of the mercantile 
 marine. 
 
 The remarks relating to the disturbance caused by 
 electric installation will, I believe, supply "a long felt 
 want" to navigating officers. 
 
 In conclusion I can only thank my friends for their 
 disinterested assistance in the shape of suggestions and 
 criticisms. 
 
 J. WHITLY DIXON. 
 
 5, Upper Montagu Street, 
 PoETitiN Square, "W. 
 August, 1898, 
 
LIST OF AUTH0KITIE8 CONSULTED IN THE 
 PKBPARATION OF THIS WORK. 
 
 Admiralty Manual of Scientific Enquiry. 
 
 Admiralty Publications on the Deviation of the Compass. 
 
 Bedford's Sailors' Pocket-Book. 
 
 Clerk-Maxwell's Electricity and Magnetism. 
 
 Collet's Compensation of Compass (translated by W. Bottomley). 
 
 Deschanel's Natural Philosophy. 
 
 Elementary Manual for the Deviation of the Compass. 
 
 Encyclopedia Britannica. 
 
 Encyclopaedia Metropolitana. 
 
 Ganot's Physics. 
 
 Jenkin's Electricity and Magnetism. 
 
 Kelvin's — Thomson's Instructions for the Adjustment of Patent Compass 
 
 Lardner's Electricity. 
 
 Lecky's Wrinkles. 
 
 Liverpool Compass Committee, Reports by. 
 
 Martin. Lectures delivered at the Royal Naval College. 
 
 Merrifield's Magnetism and Deviation of the Compass. 
 
 Raper's Navigation. 
 
 Reinold and Waghom. Deviation of the Compass. 
 
 Royal Society, Philosophical Transactions of. 
 
 Royal United Service Institution, Journals of. 
 
 Towson's Practical Information on the Deviation of the Compass. 
 
 Walker's Terrestrial and Cosmical Magnetism. 
 
ALGEBEAIC EULES. 
 
 The following simple algebraical r»les and trigonometrical ratios should 
 be known and attended to by the reader : — 
 
 Addition. 
 4- quantities added together give a + result. 
 
 )) )> )> )J 5> 
 
 To add together (+ and — ) quantities, take the difference of their sums, 
 prefixing the sign of the greater. 
 
 Subtraction. 
 Change the sign of the lower line and proceed as in addition. 
 
 Multiplication. 
 
 Like signs (+ +) (' ) give +; Unhke signs {-\ ) ( h) give 
 
 - sign. 
 
 Division. 
 
 If the Divisor and Dividend have like signs, the sign of the quotient 
 will be +, and if the signs be unlike, the sign of the quotient will be — . 
 
 To obtain the logarithm of an angle — 
 
 Between 90 ° and 180 ° stibtract angle from 180 °. 
 Between 180 ° and 270 ° subtract 180 ° from the angle. 
 Between 270 ° and 360 ° subtract angle from 360 °. 
 
 Below are the changes of sign in the trigonometrical ratios as an angle 
 
 increases from ° to 360 °. 
 
 Sin. Cob. Tan. Cot. Sec. Coseo. 
 1st Quadrant 0°to 90°. + + + + + + 
 
 2nd Quadrant 90° to 180°. + - - - - + 
 
 3rd Quadrant 180 ° to 270 ° . - - + + 
 
 4th Quadrant 270 ° to 360 ° . - + - - + 
 
CONTENTS. 
 
 CHAPTER I. 
 
 PAGE 
 
 Magnets — poles, magnetic field, magnetic fluid, magnetic substances, 
 saturation point— Soft and Hard iron .... 1 
 
 CHAPTEE II. 
 
 Terrestrial Magnetism — Variation, Dip and Intensity ; their 
 Diurnal, Annual and secular changes ..... 9 
 
 CHAPTER III. 
 
 ERRATA. 
 
 Page 1 line 8 f or " for" read " form." 
 
 >> 7 ,, 19 for "clink" read "click." 
 „ 32 „ 3 for "is "read "be." 
 
 „ 33 ,, 13 for "athwart " read "athwartahip." 
 
 „ 42 ,, 15 for "connected "read "corrected." 
 
 „ 43 „ 16 for " connected " read " corrected. " 
 
 ,,64 „ 12 for "principles "read "principle," 
 
 „ 67 (line 3 from bottom) for " pen " read "pin." 
 
 Co-EPPiciBNTS — A, B, 0, D, AND E — their meaning and values 
 ascertained — A (lambda) value of . 
 
 HebLing error — Causes of — /x (mu) value of — correction of Heeling 
 error by Vibrations and Heeling error instrument — Heeling 
 CO-EFPIOIENT 46 
 
 CHAPTER VI. 
 Compass— Mechanical correction of — General remarks and cautions . 60 
 
 Tables and Magnetic charts. 
 
ALGEBEAIO EULES. 
 
 The following simple algebraical rules and trigonometrical ratios should 
 be known and attended to by the reader : — 
 
 Addition. 
 
 + quantities added together give a + result. 
 
 "^ )? J) )j )j )j 
 
 To add together (+ and — ) quantities, take the difference of their sums, 
 prefixing the sign of the greater. 
 
 "■"'■ 
 Change the sign of the lower line and rirn^^mJ — --^ ■.-.•.■ 
 
 m^^*-^- 
 
 Below are the changes of sign in the trigonometrical ratios as an angle 
 increases from ° to 360 °. 
 
 
 Sin. 
 
 Cos. 
 
 Tan. 
 
 Cot. 
 
 Sec. 
 
 Coseo. 
 
 1st Quadrant 0°to 90°. 
 
 + 
 
 + 
 
 + 
 
 + 
 
 + 
 
 + 
 
 2nd Quadrant 90° to 180°. 
 
 + 
 
 — 
 
 — 
 
 — 
 
 — 
 
 + 
 
 3rd Quadrant 180 ° to 270 ° . 
 
 - 
 
 - 
 
 + 
 
 + 
 
 — 
 
 — 
 
 4th Quadrant 270 ° to 360 ° . 
 
 — 
 
 + 
 
 — 
 
 — 
 
 + 
 
 — 
 
CONTENTS. 
 
 CHAPTEE T. 
 
 ^ PAGE 
 
 Magnets — poles, magnetic field, magnetic fluid, magnetic substances, 
 saturation point— Soft and Haed iron .... 1 
 
 CHAPTER II. 
 
 Teeeesteial Maghietism — ^Variation, Dip and Intensity ; their 
 Diurnal, Annual and secular changes ..... 9 
 
 CHAPTER III. 
 Compasses — Lord Kbltin, Liquid, Peichl, and Landing — Beaeing 
 
 PLATE — POSITION OF COMPASS, PEDESTAL FITTINGS — ElBCTEICAL 
 
 Installation 20 
 
 CHAPTER IV. 
 
 Deviation, by distant object, terrestrial or celestial, Recipeocal 
 bearings, Dbflectoes, Conteol compass — Peemanent and 
 Sub-peemanent magnetism 
 
 Gaussin eeeoe in ship or torpedo boats — Sub-maeine disturbance . 
 
 Causes of Semi-cieculae, Quadeantal, Heeling and Constant 
 Eeeoes — Heeling Go-efficient 30 
 
 CHAPTER V. 
 
 Co-EPPiciENTS — A, B, 0, D, AND B — their meaning and values 
 ascertained — A (lambda) value of . 
 
 HeeLing eeroe — Causes of — /* (mu) value of — correction of Heeling 
 error by Vibrations and Heeling error instrument — Heeling 
 co-efficient 46 
 
 CHAPTER VI. 
 Compass— Mechanical correction of — General remarks and cautions . 60 
 
 Tables and Magnetic chaets. 
 
THE MAEINEE'S COMPASS. 
 
 CHAPTEE I. 
 
 Magnetism. — The following sentences from the works of 
 three great philosophers may fitly preface any work, how- 
 ever small or simple, relating to magnetism. The late 
 Sir John Henechel, writing on the magnetic condition of 
 our globe, states : " The relations of terrestrial magnetism 
 lie among those mysterious powers which seem to constitute 
 the chief arcana of inanimate nature and its phenomena 
 for a singular exception to the character of stability and 
 permanence which prevails in every other department of 
 the general subject." 
 
 M. Ajago writes : " Nothing in the vast domain of 
 terrestrial physics is more obscure and more uncertain, than 
 the causes which everywhere occasion the three elements of 
 terrestrial magnetism, viz., the declination, inclination and 
 intensity of the force to vary." 
 
 The late Sir Edward Sabine also writes : "All attempts 
 that have hitherto been made to connect the secular mag- 
 netic change with any other physical phenomena, either 
 terrestrial or cosmical, have signally failed." 
 
 Magnetism. — Certain substances, natural or artificial, 
 have the power of attracting iron, and are popularly known 
 as Magnets, the cause of this attraction is called magnetism. 
 
 The natural magnet or Lodestone (Fcg O4) was formerly 
 considered to possess magical qualities, and it was not until 
 the tenth or twelfth century that its property of pointing 
 
2, TEE MAMINJER'S COMPASS. 
 
 when suspended by a thread to the magnetic meridian was 
 discovered. 
 
 The Lodestone, from the Sazon " Iseden," to lead, is 
 found in the older geological formations in Sweden, Spain, 
 Arkansas, the Isle of Elba and other portions of the world. 
 
 Artificial Magnets. — Any bar of iron or steel, if 
 rubbed by a lodestone or artificial magnet will be found to 
 be magnetised, that is to say, if freely suspended it will take 
 up a known position with regard to the astronomical or true 
 north and south, and alsoi with regard to the horizon. 
 
 If the magnetised bar be placed in iron filings, it is 
 found that they are most strongly attached towards the 
 ends, decreasing towards the centre, where a neutral position 
 exists, and no filings will be attracted. 
 
 Poles. — The points neaj the end where the greatest- 
 attraction exists are called poles ; the imaginary line joining 
 them the axis of the magnet, and Dr. Gilbert, one of the 
 earliest writers on magnetism, calls the neutral position the 
 equator of the magnet. 
 
 Poles, Position of. — The poles of a magnet, are not at 
 its extremities, but generally a short distance from the end. 
 The position of poles is generally about y'^th of the whole 
 length of the bar from its ends ; the thinner a bar or needle 
 the nearer the poles approach the extremities. 
 
 The two poles in the previous experiment with the iron 
 filings would appear to be identical, but this is not so, foic. 
 when a similar pole of a magnetic needle or bar is brought 
 to that of a freely suspended magnetic needle, repulsion 
 takes place ; if the opposite pole is brought near the needle, 
 attraction takes place, hence the following rule. 
 
 Poles of the same name repel and poles of a contrary 
 name attract one another, with the undermentioned ex- 
 ception. If the red pole of a strong magnet be held at 
 some distance from the red end of a weak magnet, it will 
 repel the latter, but if brought closer, it will attract it. 
 This apparent paradox is owing to the magnetism induced 
 
MAGNETISM. 3 
 
 by the stronger magnet in the weaker, is opposite to the 
 original naagnetism of the latter and the induced magnetism 
 finally overcomes the original magnetism of the weaker 
 magnet 
 
 Earth as a Magnet. — Dr. Gilbert, towards the end of 
 the sixteenth century, discovered the inherent magnetism of 
 the earth, speaking generally, there are two poles of opposite 
 character with a magnetic equator midway between them.* , 
 
 Names of Poles. — ^Bearing in mind the law previously 
 mentioned that poles of the same name repel, and poles of 
 a contrary name attract one another, it is evident that the 
 magnetism of the north polar regions is opposite in name to 
 that existing in the pole of a compass needle which points to 
 the north ; hence the confusion in nomenckture between 
 various writers ; to avoid all ambiguity, we shall call the 
 poles of a magnet that attracts the north-seeking end of the 
 needle blue, the repelling one red, bearing in mind the pole 
 in the north end of a compass needle is a true south pole, 
 and that in the south end of a compass needle is a true 
 north pole. 
 
 When speaking of the Mariner's compass the custom of 
 seamen will be followed, calling the points according to the 
 graduation of the compass, f 
 
 The late Sir Gr. B. Airy, in diagrams, marks the magnets 
 
 111 
 
 Red magnetism 
 Blue magnetism 
 
 Makers usually mark| the red pole of a magnet by a 
 transversal cut. 
 
 * See further remarks, under the heading of Dip and Intensity, pages 
 16-19. 
 
 t It is to be regretted that all Mariner's compasses are not graduated 
 from 0° to 360. 
 
 I For marking of service magnets, see page 66. 
 
 B 2 
 
4 THE MAMINBR'S COMPASS. 
 
 Pole, Strength of. * — Is defined as proportional to the 
 force which it is capable of exerting on another given pole ; 
 hence the force (/) exerted between two poles of the 
 strengths (m) and (mi) must be proportional to the product 
 m nil. The force (/) is also found to be inversely pro- 
 portional to the square of the distance (D) separating the 
 poles, and to depend on no other quantity ; hence choosing 
 the units correctly, 
 
 /="^ (1) 
 
 The strength of a pole Is a magnitude which must be 
 measured in terms of some unit. When in the above 
 equation we make (/) and (D) both equal to unity, the 
 product m mj must also be equal to unity, hence from 
 equation (l) it follows that " the unit pole is that which at 
 the unit distance repels another similar and equal pole with 
 unit force." 
 
 The law may be briefly stated thus — the force exerted 
 between two magnetic poles is proportional to the product of 
 their strengths, and is inversely proportional to the square 
 of the distance between them : This law of inverse squares 
 is only true when the distance between the poles is suffi- 
 ciently great that the magnetism due to mutual induction 
 may be safely neglected. 
 
 The effect of a small magnet on a distant magnetic mass 
 depending on the attraction of one pole and the repulsion of 
 the other will be inversely as the cube of the distance 
 nearly. 
 
 The attraction between two magnetisms (the permanent 
 magnetism of the magnet, and the induced magnetism of 
 the iron), which is as the product of these magnetisms 
 directly, and as the square of the distance inversely ; will 
 
 * Magnetic Unit. — A Unit Magnetic Pole is one of such a strength 
 that, when placed at a distance of one centimetre from a similar pole of 
 equal strength, it repels it with a force of one dyne. 
 
MAGNETISM. 5 
 
 be inversely as tJie fourth power of tlie distance. With 
 increase of distance, therefore, the attraction diminishes 
 very rapidly. 
 
 Position of Poles when Magnet is Broken. — If a 
 magnet be broken into several pieces, each portion becomes 
 a complete magnet, with a red and blue pole, and each 
 piece is almost as strongly magnetised as the original 
 magnet, if the pieces be again brought closely together, 
 the pieces revert to a single magnet, with a red and blue 
 pole. 
 
 Magnetic Field. — The neighbourhood of the magnet is 
 often for convenience called a magnetic field, and the effect 
 produced by a magnet is often spoken of as due to the 
 magnetic field. The lines of force in a uniform magnetic 
 field will be parallel. The distance between the poles of a 
 magnet, multiplied by the force of either pole, is called the 
 Moment of the Magnet. The intensity of magnetisation 
 of a magnet is the ratio of its magnetic moment to its 
 volume. 
 
 Magnetic Futid. — In order to explain the various 
 phenomena of magnetism, two hypothetical magnetic fluids 
 are assumed (the fluids are not liquids) and represent an 
 invisible, elastic, gaseous atmosphere surrounding the 
 particles of all magnetic substances ; the fluid facing 
 the north has red magnetism, and the one to the south 
 blue.* 
 
 Before magnetisation, it is assumed, these fluids are 
 combined round each molecule and mutually neutralise 
 each other ; they can be separated by the influence of a 
 force greater than that of their mutual attraction, and caji 
 arrange themselves round the molecules, to which they are 
 attached, but cannot be removed from them. 
 
 The above hypothesis appears to be convenient, more 
 especially with regard to that magnetic phenomena which 
 
 * The French writers call the fluid facing the north Boreal, and the 
 other Austral. 
 
6 THE MABINER8 COMPASS. 
 
 appear to result from electrical currents, circulating in 
 magnetic bodies, connecting tlie theory of magnetism with 
 that of electricity. 
 
 Coercive Force. — Soft-iron * is either malleable iron 
 which has not been hammered or subjected to any violence 
 when cold ; or cast iron. It is instantly magnetised by 
 induction when exposed to any magnetic force, but has no 
 power of retaining magnetism.f Hard iron* on the contrary, 
 when once magnetised retains its magnetism, when the 
 inducing force is removed — hard iron or steel may be 
 magnetised by percussion, if held vertically, or still better in 
 the line of dip. A hard iron bar or steel may be artificially 
 converted into a magnet by means of a single powerful 
 magnet drawn from fifteen to twenty times in the same 
 direction along the bar or needle ; the end last touched 
 being of opposite polarity to the pole of the magnet used ; 
 this method should only be employed for small magnets or 
 needles. The usual process is to apply to the centre of a 
 steel bar when laid on wood, the opposite poles of two 
 magnets held obliquely at an angle' of about 30° to the steel 
 bar, drawing the magnets about ten times slowly along to 
 the opposite ends of the bar, then turn the bar over and 
 treat it ' in a similar manner on the other face ; it will then 
 be found that the steel bar has become permanently 
 magnetised. 
 
 A third method of producing most powerful magnets is 
 by rods or bundles of wrought iron round which an electric 
 current circulates in a coil of wire — in a steel magnet the 
 permanency depends on the quality of the steel; for 
 
 * The reader should clearly understand that the iron used in the 
 construction of a ship there are many descriptions that cannot be classed 
 as either hard or soft iron. No iron is perfectly soft, and no steel perfectly 
 hard. Theoretically the magnetism in a magnet is not quite fixed, for no 
 steel is so hard as not to be temporarily affected by magnetic induction. 
 
 t A piece of soft iron, which is i a magnet by induction, can induce 
 magnetism in another piece of soft iron ; thus a magnet may sustain a 
 long string of nails, each hanging to the one above it. 
 
MAGNETISM. 7 
 
 compass needles tHe steel should be perfectly uniform in 
 hardness but of spring temper. 
 
 Surface Magnetisation. — In many cases the magnetism 
 imparted to magnets is confined chiefly to the outer layers of 
 steel, if the outer layers are dissolved by acid, it loses its 
 magnetism after a thin film has been removed. Magnets, 
 which have been thoroughly magnetised, however, exhibit 
 some magnetism in the interior. A hollow steel tube when 
 magnetised is nearly as strong a magnet as a solid rod of 
 the same size : if a bundle of bound steel plates are 
 magnetised, the inner ones may occasionally exhibit a 
 reversed magnetisation. Heat diminishes the power of a 
 magnet, if a steel one is raised to a white heat it is 
 permanently demagnetised. 
 
 Mechanical Effects of Magnetisation. — Joule found 
 an iron bar to increase by t^qooq of its length when 
 powerfully magnetised, but as its volume remained the same, 
 it contracted in thickness — if magnetisation or demagnetisa- 
 tion is effected suddenly, a faint metallic clink is heard. A 
 twisted iron wire tends to untwist itself when magnetised : a 
 piece of iron when powerfully magnetised and demagnetised 
 in rapid succession grows hot. 
 
 Magnetic Substances. — Iron, nickel and cobalt are 
 (magnetic) * paramagnetic, being attracted to either pole of 
 the magnet, and if suspended between the poles of a magnet 
 tends axially, that is, to place its length parallel to the lines 
 of force. 
 
 Bismuth and antimony are diamagnetic bodies, being 
 repelled by either pole of a magnet, and if suspended 
 between the poles of a magnet, tend equatorially , that is, to 
 place its length at right angles to the line of force. 
 
 Saturation Point. — When the highest degree of retain- 
 able magnetisation is reached in a steel bar, it is said to be 
 magnetised to saturation ; by means of powerful magnets, a 
 
 * Strictly, the word magnetic onglit to be a general one, including all 
 the phenomena and effects produced by that power. 
 
« TEE MABINERS COMPASS. 
 
 steel bar, more especially if newly made, may be temporarily 
 magnetised beyond saturation, but it rapidly returns to 
 saturation point. 
 
 The Force in a magnetised needle is simply directive, but 
 has no power to move the needle, is evidenced by the fact 
 that a magnetised needle floating freely on cork in a liquid, 
 does not move, the attra,cting force of the red pole is exactly 
 equalled by the repelling force of the blue ; in addition the 
 weight of a needle is the same before and after magneti- 
 sation. 
 
TEBRESTRIAL MAGNETISM. 
 
 CHAPTER II. 
 
 Terrestrial Magnetism. — -A freely suspended magnetic 
 needle takes up a certain position with, regard to the true or 
 geographical north, and also with regard to the horizon, the 
 phenomena relating thereto are called terrestrial magnetism ; 
 the three elements required in order to denote the mag- 
 netism of any place on the earth's surface are the Variation 
 (declination), • Dip (inclination), and / the Intensity of the 
 Magnetic Force. The two former are constantly brought 
 under the notice of a seaman during a voyage, and the third, 
 Intensity, is the force which tends to set itself in the direc- 
 tion of the line of the dipping needle, that is in the lines of 
 force, but for purposes of compass correction, it is better to 
 consider this as being composed of two component parts — a 
 horizontal force acting parallel to the earth's surface, and the 
 other a vertical one, acting at right angles to the horizontal 
 force, pulling the red end of the needle downwards in the 
 northern magnetic hemisphere, and upwards in the southern 
 one. 
 
 Lines of equal Variation, Dip, and Intensity, are called 
 IsoGONic, IsocLiNic, and Isodynamic lines respectively. 
 
 The term Magnetic Eqqator is usually applied to the 
 line of no dip, but it would be more correctly applied to the 
 curve which cuts all the magnetic meridians at right angles. 
 The Dynamic Equator is the curve on the earth's surface 
 connecting the points of least magnetic intensity. Mag- 
 netic Poles are positions of vertical dip (90°), but they do 
 not coincide with those of the greatest vertical force ; also 
 the positions of greatest intensity do not coincide with the 
 
10 THE MARINER'S COMPASS. 
 
 magnetic poles, the former are usually called magnetic foci 
 or centres. The above magnetic elements are subject to 
 regular and irregular changes, the former are called Secular, 
 Diurnal, and Annual ; and to the irregular changes, when 
 taking place simultaneously over large surfaces of the earth, 
 the term Magnetic storms have been applied. 
 
 Various hypotheses have been advanced to connect the 
 above elements obtained by observation with a system, but 
 although the general facts naay be shown, there are distinct 
 discrepancies between theory and observation, and it is 
 probable that the earth's magnetism cannot be reduced to 
 any mathematical expression. Halley, Mayer, Humboldt, 
 Biot, Hansteen and Gauss, have all advanced theories on the 
 subject. 
 
 The cause of the \earth's magnetism remains as yet un- 
 explained. The late Sir G. B. Airy considered that terres- 
 trial magnetism was not produced to any large extent in the 
 earth's surface, supporting the latter proposition by observa- 
 tions made on mountains and in balloons.* Gauss considered 
 that the agents producing the magnetic force of the earth, 
 or the greater portion of them, were situated exclusively in 
 the interior of the earth ; Gauss also considered that the cause 
 of terrestrial magnetism was not external to the earth. The 
 late Sir G. B. Airy states : " There is reason to believe that 
 the sun and moon act as magnets," this has been con- 
 clusively proved in the ' Adams' Prize Essay for 1865,' by 
 E. "Walker, M.A.f Magnetic disturbances occur simul- 
 taneously at places thousands of miles apart, their periodical 
 nature, length, times of maxima and minima, being coinci- 
 dent with those of solar spots ; and that the magnetic force in 
 
 * This question is still further complicated by the fact that the amount 
 of oxygen in any known volume of air will vary with its altitude, and 
 oxygen is paramagnetic. 
 
 t Published by Deighton, BeU & Co., Cambridge. This excellent 
 work, for clearness and conciseness, leaves nothing to be desired, and 
 should be in the hands of every naval officer interested in magnetism. 
 Chapter II. is largely based on it. 
 
VACATION. 11 
 
 both hemispheres during the months of December, January 
 and February, is greater when the sun is nearest the earth ; 
 than in May, June and -July, when it is more distant. The 
 discovery of a semi-annual * and lunar-diurnal change in the 
 diurnal variation, point to a cosmical origin, but whether the 
 sun acts by direct heating power (which appears unlikely), 
 or indirectly ; and the moon from induced magnetism received 
 from the earth, is at present undetermined. It is probable 
 that the agents producing terrestrial magnetism are both 
 physical and cosmical, modified or complicated by earth- 
 quakes, volcanic eruptions, and sudden atmospheric change6,f 
 and the secret of terrestrial magnetism is " reserved for the 
 industry of future ages." 
 
 Vaeiation (DBCLiisrATiON) is the horizontal angle between 
 the true and magnetic meridians. The first European to 
 note the variation of the compass needle was Columbus, on 
 September 13th, 1492, in latitude 28° N., and longitude 
 28° W. ; doubtless the early navigators knew that the true 
 and magnetic meridians did not coincide, but Columbus 
 found that the Variation itself changed ; in China, the 
 Variation was mentioned in th& early part of the twelfth 
 century. Boroughs, Comptroller to the Navy in Elizabeth's 
 reign, published a work called * A Discourse on the Varia- 
 tion of the Cumpas or Magneticall Needle ' ; the variation at 
 Limehouse (16th October, 1580) was 11° 18' E. The secular 
 change in the Variation was made known by Professor Gelli- 
 
 * See pages 13 and 14. 
 
 ■j" Wind-storms do not appear to exercise any perceptible influence on 
 the magnetic needle, but Grilpin states, from observations made in 1786-87 
 at Somerset House, that he found a change of wind produced a variation 
 in the needle from steady to unsteady, or the contrary. The needle was 
 most, steady when the wind was south or south-west, and most unsteady 
 when it was east. — Walker, page 63. With regard to thunderstorms, the 
 late Sir F. Evans (apparently on the authority of the late Sir G. B. Airy) 
 states that they exercised no perceptible influence on the needle. Mac- 
 donald in 1794, and the late Father Secchi held contrary opinions. With 
 r^ard to lightning, in wood-built vessels the poles of the needles have 
 been reversed ; in an iron-built ship, if struck with lightning, her magnetic 
 character would probably undergo considerable change. 
 
12 
 
 THE MARINER'S COMPASS. 
 
 brand, of Gresham College, in a work called ' A Discourse 
 Mathematicall on the Variation of the Magneticall Needle, 
 together with its admirable diminution lately discovered ' ; 
 from this date, the secular change in the Variation was 
 recognised. The following tables showing the secular 
 change in the Variation for London and Cape of Good Hope 
 
 London. 
 
 Easterly variation marked + 
 Westerly variation marked — 
 
 Cape of Good Hope. 
 
 Annual eastward change + 
 Annual westward change — 
 
 Date. 
 
 Variation. 
 
 Mean Annual 
 Westward cbange. 
 
 Date. 
 
 Variation. 
 
 Meun Annual 
 Westward change. 
 
 1580. , 
 
 11-15 + 
 
 
 1605. . 
 
 0-30 + 
 
 . , 
 
 1634. . 
 
 4-6 + 
 
 10-6'- 
 
 1675. . 
 
 8-14- 
 
 7-06- 
 
 1672. . 
 
 2-30- 
 
 ,10-5'- 
 
 1751. . 
 
 19-15- 
 
 6-62- 
 
 1723. . 
 
 1773. . 
 1805. . 
 
 14-17- 
 21-9- 
 
 24-8-, 
 
 8-1- 
 9-3- 
 0-7- 
 
 1792. . 
 1839. . 
 1842. . 
 
 24-31- 
 29-09- 
 29-06- 
 
 5-68- 
 J 13-00- 
 t (in 1836) 
 
 0-29 + 
 
 1818. . 
 
 24-43- 
 
 Maximum 
 
 1849. . 
 
 29-16- 
 
 2-15- 
 
 1858. . 
 
 21-54- 
 
 6-8 + 
 
 1895. . 
 
 29-24- 
 
 
 1898. . 
 
 16-59- 
 
 7-0 + 
 
 1898. . 
 
 29-16- 
 
 2-5+ 
 
 The Diurnal change in Variation was first noticed by 
 Mr. Graham, an instrument maker of London, in 1722, 
 finding that the diuxnal variation being about 50' (from 
 14° -45' to 13°" 50'). In a paper read before the Eoyal 
 Society, by Canton, on December 13th, 1759, it was shown 
 that the diurnal change was generally regular, i.e., the varia- 
 tion of the needle westward increased from 8 or 9 a.m. until 
 1 or 2 P.M. ; the needle remained stationary for some 
 time, but proceeding to return to its approximate normal 
 position during the night or by next morning. Canton 
 found that the extent of the diurnal range was greater in 
 
VARIATION. 13 
 
 summer tlian in winter.* Macdonald, in Sumatra, in 1794, 
 found that the easterly diurnal change of the variation in- 
 creased during from 7 a.m. to 5 p.m., and decreased from 
 that time until 7 A.M. the next day. From his observa- 
 tions,! Macdonald deduced two general rules — 
 
 (1.) That the amount of the diurnal change is less in the 
 tropics than in temperate climates. 
 
 (2.) The motion of the needle in the southern hemi- 
 sphere is in the opposite direction to that in 
 which it moves in the northern hemisphere at 
 the same hour. 
 
 The first rule ia generally correct, but no definite relation 
 exists between the latitude and the extent of the daily 
 change. The second rule is not correct so far as tropical 
 regions are concerned. The late General Sabine, from five 
 years' observations made at Singapore, St. Helena, and the 
 Cape of Good Hope, shows " that, for one half of the year, 
 the diurnal motion of the needle corresponds to that observed 
 in the northern hemisphere at the same hours, whilst for 
 the other half it corresponds to the motion in the southern 
 hemisphere." The passage from one direction of motion 
 to the other takes place soon after the equinoxes, i.e., when 
 the sun crosses the terrestrial equator ; at this period of the 
 year, the diurnal variation partakes more or less on difierent 
 days of the character of both seasons, but at all other periods 
 the contrariety is unmistakable, " the north end of the needle 
 reaching its eastern extreme in one case, and its western 
 extreme in the other, at nearly the same hours, and it must 
 be further noticed, that these extremes are, in both cases, 
 
 * Canton also noticed that irregular movements of the needle were 
 accompanied by an Aurora BoreaKs, a fact also discovered by Wargenten 
 in 1750. 
 
 t During these observations Macdonald inferred a diurnal change in 
 the dip, noticing that the needle did not always remain at the same 
 inclination to the horizontal plane. 
 
14 TEE MARINER'S COMPASS. 
 
 equidistant from the mean position of the needle in the 
 respective months." 
 
 Investigation of the above observations revealed the 
 existence of a semi-annual change in the diurnal change, 
 having opposite phases, according as the sun's dfeclination is 
 north and south. 
 
 LuNAR-DiTJRNAL Changb. — The influence of the moon on 
 the three magnetic elements was first announced by M. KreU 
 in 1841, and confirmed by General Sabine in 1856, stated 
 briefly, the variation has two easterly and two westerly 
 maxima, and the dip and total force have likewise each two 
 maxima and two minima in the same interval, the change in 
 each case passing through zero four times during the lunar 
 day.* 
 
 In Paris, the diurnal change of variation reaches its 
 maximum eastward direction at %\ h. A.M., 8|- p.m., and 
 maximum westward direction at 1\ p.m. and 11 p.m., 
 agreeing closely with those at Kew. The diurnal change 
 varies in different months, the greatest range taking 
 place in April (13'), and the least in December (6'). 
 The change in the position of the needle is effected 
 not by a steady progression, but by a series of small 
 oscillations. 
 
 Astatic Needle. — If it be required to ascertain the 
 maximum and minimum effect of the disturbing cause on 
 the magnetic needle, and not the actual amount of the 
 variation, the needle is made astatic by placing two needles 
 of the same force parallel to each other, with their poles in a 
 contrary direction, but with the two needles rigidly con- 
 nected at their centres ; the terrestrial magnetic direction is 
 neutralised, the needle retaining all its natural power, and 
 its sensitiveness is greatly increased. 
 
 * The lunar magnetic influence is less felt in winter, owing to the 
 effect being so smaU, that it is probably swallowed up in the greater 
 irregular disturbances that prevail with more frequency during the winter 
 than in summer. 
 
VARIATION. 15 
 
 Annual or Periodical change of the Variation. — In 
 addition to tlie secular change, it was discovered by Cassini, 
 in 1786, that the magnetic, needle was subject to small 
 annual periodical fluctuations. From the observations of 
 Ca^ini and Gilpin, M. Arago suggested certain hypotheses 
 with regard to the annual cha,nge, which subsequent ex- 
 perience has not confirmed; they are, therefore, not men- 
 tioned.* It is almost certain there does exist on the globe 
 an annual change of the variation, irrespective of the direc- 
 tion or magnitude of the variation, its secular change, or the 
 geographical position of the observer ; and depending on the 
 sun's position with reference to the equator, causing ^e red 
 end of the needle to point more to the east when the sun is 
 north, and to the west when it is south of the equator ; the 
 turning-points must, tjierefore, be approximately at the 
 times of the equinoxes. The amount of the annual change 
 is small, and varies, as stated above, at different times of 
 the year. 
 
 Irregular changes in the Variation occur during earth- 
 quakes, volcanic eruptions, and the appearance of the Aurora ; 
 the effect of the latter on the magnetic needle extends con- 
 siderably beyond its visibility. The magnetic perturbations 
 or magnetic storms have been simultaneously observed at 
 Toronto, Cape of Good Hope, Prague, and Hobart, and 
 have been connected by the late General Sabine with the 
 periods of maximum sun spots, i.e., approximately every 
 eleven years. 
 
 The variation of the compass in shore observatories is 
 obtained by means of the declination compass,f if desirous 
 of ascertaining the variation, by the compass supplied to 
 Her Majesty's ships (landing one). J Obtain the true 
 bearing of six well-defined objects, about 60° apart (to 
 eliminate, as far as possible, errors of centring, graduation, 
 etc.), from an observation spot free from iron ; the mean 
 
 * Walker, pages 63, 70, 72, 76. f Ganot's ' Physics,' page 67. 
 
 X See page 23. 
 
16 THE MAMINEB'8 COMPASS. 
 
 of the bearings of each object, by attracting and repelling 
 the red end of the needle, should be used. 
 
 Different caps and pivots should be employed to prevent 
 error in this respect ; the observations should be repeated at 
 such hours that the variation wUl not be affected by the 
 diurnal change ; the error of the observing card, if known, 
 should be applied. 
 
 Dip or Inclination * was first observed by Norman, an 
 instrument maker in 1756 ; it is the angular difference 
 measured in the vertical plane, between the direction of the 
 needle and the horizontal plane or zero. If the red end of 
 needle dips below the horizon, it is usually called north dip, 
 with a + sign afl&xed ; when the blue end of the needle dips 
 below the horizon it is called south dip, with a — sign 
 affixed.f In a vessel proceeding from London with a dip of 
 + 67° to the Cape of Good Hope with a — 58°, an observer 
 would notice the needle gradually assuming a horizontal 
 position ; on reaching the magnetic equator (Aclinic line), 
 the dip would be nil and the needle horizontal ; proceeding 
 south, the blue end of the needle would gradually dip, until, 
 on arrival at the Cape of Good Hope, it would make an angle 
 of 58° with the horizontal plane. 
 
 The dip is subject to secular, diurnal, and annual changes 
 in a similar manner to the variation. The secular change 
 varies in different portions of the globe. The Bight of Benin 
 is where one of the most rapid changes takes place ; at 
 Tahiti, the change in sixty-seven years was 35' ; at the 
 present time (1898), the annual change at the Cape of Good 
 Hope and Kew is respectively 5' increasing, and l'"7 de- 
 creasing. 
 
 The diurnal change of the dip has its regular horary 
 variations, turning points, and its maxima and minima, they 
 
 * The relationship between dip, horizontal, vertical and total force will 
 be shown under the heading' of Intensity. See page 18. 
 
 t The method of obtaining the dip both on shore and at sea may be 
 studied in the ' Admiralty Manual of Scientiiic Enquiry,' article ' Terrestrial 
 Magnetism ; ' Walker, page 155 ; and Ganot's ' Physics,' page 581. 
 
DIP. 
 
 17 
 
 are as well defined as those of the variation, but they vary 
 at different geographical positions. At some observing 
 stations a semi-annual period has been discovered in the 
 hours of the principal maxima and minima, viz., from April 
 to September, and from October to March ; but as the diurnal 
 change in the dip is only a few minutes, it has no practical 
 value to a seaman. 
 
 Semi-annual Change of Dip. — From observations made 
 at Toronto with a + dip, it appears that the inclination is 
 less during the months when the sun is south of the 
 equator, than when it is north of it, the mean result 
 coinciding approximately to the time of the equinoxes ; 
 a similar result is found from Hobart observations with 
 a — dip, the amount being about 15'. 
 
 The following table shows the secular change in the dip 
 for the neighbourhood of London *: — 
 
 Year. 
 
 1580 
 1672 
 1723 
 1773 
 1821 
 1860 
 1898 
 
 Dip. 
 
 71 50 + 
 
 73 30+ 
 
 74 42+ 
 
 72 19 + 
 70 03 + 
 68 19 + 
 67 20 + 
 
 Intensity. — Havmg given a general description of 
 terrestrial magnetism so far as it relates to direction, we 
 now consider its strength or intensity. To Humboldt is given 
 the credit to demonstrate its existence in 1798-1803. As 
 previously mentioned, the total force acts in the line of dip ; 
 it is generally resolved into its compQnent parts, horizontal 
 force and vertical force, and for the purpose of this work, 
 horizontal force and vertical force, dip, and the variation are 
 
 * Walker, page 177. 
 
 G 
 
18 
 
 TEE MARINER'S COMPASS. 
 
 considered. The connexion shown in the figure below is for 
 Kew and the south coast of England generally. 
 
 H. F. is the horizontal force. 
 V. F. „ vertical force. 
 T. F. „ total force. 
 D. (67° 20') is the dip (red end). 
 
 H.F 
 
 -1. 
 
 >■ 
 
 V^ 
 
 
 
 
 /v 
 
 1 
 
 k 
 
 / 
 
 f 
 
 '- 
 
 ^ 
 
 H.F. 
 H.F. 
 T. F. 
 T. F. 
 V. F. 
 
 X sec D. 
 X tan D. 
 X cos D. 
 X sin D. 
 X cosec D. 
 
 = T.F. 
 = V. F. 
 = H. F. 
 = T.F. 
 = T.F. 
 
 Fig. 3. 
 
 The horizontal force is assumed to be unity for con- 
 venience, but the earth's total force, if expressed in British 
 units, varies from 6"0 to 15 "2, the units being a second of 
 time, a foot in length, and a grain in weight ; in Continental 
 countries they are a second of time, the millimetre (0 ■03937' 
 of an inch, or 0" 00328 of a foot) and the milligramme 
 (O* 01543 of a grain). To convert British into Continental 
 measure, multiply by 0*46108, and vice versd, multiply by 
 2-1688.* 
 
 The C. G. S. system is usually adopted in modern text- 
 books, it has been recommended by the Eoyal Society ; to 
 obtain the earth force according to the C. G. S. system from 
 British values, multiply the latter by 0' 04611, and divide 
 the Continental by 10. 
 
 From the above figure it will be seen that at the 
 magnetic equator the horizontal and total forces are equal, 
 the former decreasing and the latter increasing until one of 
 the magnetic poles is reached, where the horizontal force is 
 nil and the total force at a maximum. ; but the foci of 
 greatest intensity are not coincident with the magnetic 
 poles. As the total force cannot be directly measured, 
 its horizontal component is determined by vibrations of a 
 
 * ' Admiralty Manual,' article ' Terrestrial Magnetism.' 
 
INTENSITY. 19 
 
 magnetic needle in a given time, or by statical measure- 
 ments.* 
 
 Tlie Total force or Intensity is subject to secular, diurnal, 
 and annual, or semi-annual changes in a similar manner to 
 the Variation, the annual change being governed by the 
 position of the sun in the equator, being greater in both 
 hemispheres in December, January, and February, when the 
 sun is nearer the earth, than in May, June, and July, when 
 it is farther away. Gauss considered the total magnetic 
 action of the earth is the same as that which would be 
 exerted if in each cubic yard there were eight bar magnets 
 each weighing a pound. 
 
 * ' Admiralty Manual,' article ' Terrestrial Magnetism.' 
 
20 THE MARINER'S COMPASS. 
 
 CHAPTER HI. 
 
 The previous chapters treating of the laws of magnetism 
 generally, with reference to those governing the magnetic 
 needle, are intended to enable the reader more clearly to 
 understand the general laws before proceeding to describe 
 the compass itself. 
 
 The Compasses used in H.M. Navy are Sir William 
 Thomson's (Lord Kelvin), an improved Liquid Compass, 
 Peichl, and the Landing Compass, formerly the Admiralty 
 standard ; on the principal steamship companies Sir William 
 Thomson's is largely used. 
 
 Thomson's Compass. — In H.M. ships larger than the 
 "Destroyer class," Sir William Thomson's compass is used 
 as a standard, also principally in other steering positions, the 
 exceptions being the conning tower and protected stations 
 between decks.* Its method of suspension, although 
 greatly superior to the old Admiralty system, still leaves 
 room for improvement in modern fast cruisers, where the 
 vibration is considerable when steaming at high rates of 
 speed, and where it is placed on a superstructure (owing to 
 the exigencies of the service) such as a chart room. 
 
 The card is very light, possessing small needles of 
 sufficient directive power; the oscillation is slight, the greater 
 portion of the weight of the card being near its edge, thus 
 producing a long period of vibration. The outer graduation 
 of the numerals in the standard compass card is inverted to 
 enable the card to be read direct with the azimuth mirror. 
 
 * ■ See page 22. 
 
COMPASSES. 21 
 
 The compass bowl is fitted with a graduated verge for 
 horizontal angles. 
 
 Azimuth Mirror.— A description of its construction 
 would be out of place in this small work ; it is placed on the 
 upper surface of the glass in the bowl, turns freely in 
 azimuth, and is capable of observing terrestrial, solar and 
 stellar objects ; it is also fitted with a shadow pin for rough 
 observations. Care should be exercised not to communicate 
 the motion of the azimuth mirror to the bowl, and if the 
 observed object be indistinct, it is advisable to reflect the 
 card to the object. 
 
 The Liquid Compass,* used in destroyers and torpedo- 
 boats, has been found to be fairly sensitive and steady, 
 considering the class of vessel in which they are used. The 
 cards are fitted with needles, of great directive power, 
 placed edgeways, floating in a fluid composed of two portions 
 of distilled water, and one of pure spirits of wine. The 
 cap is a sapphire with a pivot point of metal (Iridium), 
 the pressure of the card on the latter should not exceed 
 one hundred grains. A means of filling is provided should 
 bubbles appear, and in addition the bowl is fitted with an 
 elastic or expansion chamber, by which the fluid passes into 
 or out of the bowl on change of temperature. 
 
 The air is exhausted (as far as practicable) from the fluid 
 by an air-pump, the bowl is then closed and hermetically 
 sealed ; should large bubbles appear it would point to a 
 serious defect, and it is desirable in such cases to return the 
 compasses to a dockyard or the makers. The liquid compass 
 when used as a standard, in addition to a shadow pin, is 
 fitted with at removable azimuth circle, carrying a sight vane 
 and a reading prism, and it is well adapted for all observa- 
 tions during daylight ;f the numerals marking the graduation 
 might be slightly increased in size with advantage. 
 
 * By E. Dent & Co., 61, Strand, London. 
 
 t A system capable of taking observations at nigbt, under the con- 
 ditions that prevail on board a destroyer or torpedo-boat at sea, is much 
 required. See pages 28 and 29. 
 
22 THE MARINERS COMPASS. 
 
 The Peichl compass is the invention of an officer of 
 tlie Austro-Hungarian Navy of that name. In protected 
 stations, such as in conning towers and between decks, the 
 quadrantal error* for a 10" Thomson's compass amounts to 
 from 16° to 26° requiring 18" to 24" spheres for its correc- 
 tion, the size of the spheres almost prohibit their use in the 
 confined space of a conning tower, the difficulty with regard 
 to space might be lessened by using a 6" compass, but the 
 remaining difficulty of the low directive force of any 
 compass in such a position would not be removed. 
 
 In the corrector f proposed by Chevalier Peichl with soft 
 iron rods, the directive force is increased, spheres and 
 Flinders bar are not used, the compass with its corrector 
 is reduced to moderate and serviceable limits : the draw- 
 backs, although they hardly deserve that name, are the 
 adjustment of the corrector on change of latitude, and unless 
 the suspension be perfect, the introduction of a heeling error. 
 The corrector consists of two metal discs, placed one above 
 the other, containing thirty-two soft iron bars movable in 
 opposite directions around the compass bowl, the inner ends 
 of which are from 1 to Ig inches in distance from the ends 
 of the compass needles and form an ellipse, whilst their 
 outer ends form a circle, the proximity of the needles to the 
 correctors induce magnetism on the latter, and as previously 
 mentioned, this necessitates an adjustment of the corrector 
 on change of latitude, the adjustment is provided for in a 
 simple manner by an index and scale. The moderate space 
 occupied by the Peichl compass and its corrector, and the 
 additional directive force of the former, point to it being a 
 suitable one for conning towers and protected positions 
 between decks, more especially in those portions of the globe 
 where the horizontal force is small. 
 
 * For causes of quadrantal error, see page 41. 
 
 t The Peichl compass is not described, as it possesses no distinctive 
 features, and a small liquid compass, such as already mentioned, appears 
 preferable. The corrector is fully described in a lecture delivered by Staff- 
 Commander (now Captain) E. Creak, Superintendent of Compasses at the 
 United Service Institution, in 1889. 
 
POSITION OF COMPASS. 23 
 
 The Admiralty Compass, now used as a landing one, 
 is well suited for that purpose being an efficient shore 
 compass. 
 
 Position of Standard Compass. — Whatever description 
 of compass is employed, careful consideration should be 
 given to the position of the standard compass. In ships of 
 the Eoyal Navy it is selected by the Controller's department 
 after consultation with the Hydrographic, due regard being 
 paid to the fact that the ship is a fighting one. In vessels 
 larger than second class cruisers, it is generally placed 
 forward, for convenience when in pilotage waters and 
 manoeuvring * generally ; it should be easily and rapidly 
 accessible by day and night,, sufiiciently elevated to obtain 
 allround views, but not placed above slight superstructures, 
 owing to the increased vibrations when steaming at high 
 rates of speed, and the increased, shocks on firing heavy 
 guns, the rules issued by the Controller's department to 
 Constructors in H.M. Navy are as follows : — 
 
 1. In all designs for the construction of iron ships, a 
 place to be prepared for the Standard compass,v and to be 
 shown in the plans. It should be placed near that part of 
 the ship from which she is navigated, and the view there- 
 from as free from obstructions as possible. 
 
 2. No iron to be placed within 5 feet of it ; 7 feet, is 
 better. 
 
 3. To be as far as possible from the ends of elongated 
 masses of iron, especially if vertical, such as funnels, 
 ventilators, masts and conning towers. 
 
 4. No iron, subject to occasional removal, should be 
 placed near the compass so as to aff'ect it. 
 
 5. In all electrical installations f care should be taken 
 that no disturbing effect be produced on the standard 
 compass. 
 
 * In many ships a manoeuvring compass is provided in addition to the 
 standard. 
 
 I Electrical installation, see piigu ^.'5. 
 
24 THE MARINERS COMPASS. 
 
 6. The awnings should be so arranged as to leave the 
 standard compass clear and available for observing the 
 deviation at all times. 
 
 Pole or mast compasses, have been found very useful 
 in cargo-carrying steamships ; they are inconvenient for 
 consultation, and the vessel's vibration and motion are 
 increased in a compass so placed. 
 
 Pelorus or Bearing plate. — Where an allround view- 
 is unobtainable, the standard compass is supplemented by a 
 Bearing Plate or Dumb Card, those supplied to H.M. ships 
 may be briefly described as follows. A gimbal ring is 
 mounted in gimbals to two short uprights secured to side of 
 its box. The bowl is weighted to keep it horizontal in a 
 seaway, and is balanced by the gimbals to the outer gimbal 
 ring. The upper surface of the bowl consists of two brass 
 plates movable in azimuth by means of raised metal studs. 
 The outer brass plate is marked with lubber lines, carrying 
 sight vane with its pointer, indicating observed bearing ; 
 the inner plate is marked to cardinal and inter-cardinal 
 points, and' is graduated to degrees. To take a bearing, set 
 inner plate by the lubber's line to ship's course being steered ; 
 observe bearing in the usual manner, and the pointer will 
 denote the observed bearing on the graduated inner plate; 
 observe simultaneously when the ship is on her course, this 
 can be assured by placing an observer at the standard 
 compass. The lubber's line must be parallel to the centre 
 fore and aft line of the ship, i.e., it should coincide with an 
 object forward equidistant from the centre fore and aft line 
 of the ship, as the bearing plate is ; and the bowl as far as 
 practicable being kept horizontal. 
 
 The bearing plate can be utilised in a similar manner for 
 obtaining the deviation of a compass between decks. 
 
 Mechanical errors in compass card construction, are 
 principally the magnetic axis of the needles on the card are 
 not parallel with the north and south points engraved on 
 the cards, distortion of the card from any cause ; prism error. 
 
PEDESTAL FITTINGS. 25 
 
 and the lubber line not in or parallel to the fore and aft line. 
 It should also be borne in mind that pivots become blunted, 
 cups cracked, also pivots are occasionally not sufficiently 
 inserted into their holders to prevent friction ensuing 
 between the upper surface of the card and the under one of 
 the glass verge, the above defects often give rise to reports 
 of a magnetic change in a ship which are really due to 
 mechanical causes. 
 
 Pedestal Fittings. — The wooden pedestal, supporting 
 the compass bowls, the latter are usually of brass, are 
 perforated with a series of holes, fore and aft, and athwart- 
 ships, forming receptacles for correcting " the semi-circular 
 error ; * the holes should be horizontal. 
 
 The Quadrantal error f is generally corrected by a pair of 
 non-magnetic hollow iron globes, mounted on metal brackets, 
 affixed to the wooden pedestal. Cylinders with spherical 
 ends, in pairs or single, soft iron chain and thin laminae have 
 been used. Spheres are more symmetrical in their action 
 and their effect can be more easily calculated. 
 
 Through the centre of the pedestal is a vertical hole or 
 well ; in it is hung by means of a metal chain,J to enable it 
 to be lowered or raised as required, a small metal can or 
 bucket containing usually seven compartments, to prevent 
 the magnets touching one another, in these the vertical 
 magnets for correcting the Heeling error § are placed. 
 
 Dynamo. — A compass, if within the magnetic field of a 
 dynamo will be disturbed, the error varying with change of 
 azimuth. No dynamo should be placed within 35 feet of a 
 Standard compass ; with the type of dynamo in general use 
 in H.M. ships, designed for 80 volts at the terminals, the 
 
 * See Semi-Circular error, page 40. An improvement would be to 
 place the magnets in a rack, capable of being moved by the observer by 
 means of a screw. 
 
 t See Quadrantal error, page 41. 
 
 X Should the chain part it is diflBcult to recover the can ; an improve- 
 ment would be to place the latter in a rack. 
 
 § See Heeling error, page 43. 
 
26 THE MARINER'S COMPASS. 
 
 minimum distance of a compass should be 60 feet from a 
 300 ampere machine, increased to 70 feet from a 400 ampere 
 one; for 600 ampere machines (being armour-clad) the above 
 rules do not apply. A dynamo exerts the least disturbing 
 influence on a compass, when the centre line of its armature 
 produced, intersects the vertical plane passing through the 
 centre of the compass ; if the standard compass can be 
 placed at a safe position from the dynamo, this position is 
 not practically carried out ; the usual one for a dynamo is 
 with the centre line of the armature shaft either fore and aft 
 or athwartships ; the rules given above -are as far as prac- 
 ticable carried out with regard to compasses other than the 
 standard. 
 
 Having regard to the necessary arrangement of the 
 fittings in a man-of-war, the difiiculty in preventing the 
 disturbance of the compass by a dynamo has been consider- 
 ably lessened by the introduction of an armour-clad dynamo, 
 so named, because it is enclosed in a shield (f ' plate) of soft 
 iron, which so to speak, gathers the lines of magnetic force 
 into itself, almost destroying the external magnetic field. 
 All the 600 ampere machines are armour-clad, they practi- 
 cally produce no disturbance on a compass 15 feet away, 
 20 feet may therefore be considered a safe distance, but 
 furt*her experiments will be required before a definite 
 distance can authoritatively be given. 
 
 Where the ship's internal fittings compel a departure 
 from the general rules given above, and a disturbing 
 element on the compass is introduced by the proximity 
 of the dynamo, for instance (1) in "Destroyers " where the 
 dynamo (100 amperes) is placed well forward. (2) In the 
 second class cruisers (" Apollo " class), where the two 
 dynamos (300 amperes) are placed one in the after end of 
 each engine room, the standard compass being on the after 
 bridge, (l) The correction is made by an electro-magnet 
 at the foot of the compass pedestal, with its poles opposed 
 to the poles of the dynamo, the ends of the coils of each 
 
DISTURBANCE BY DYNAMO. 27 
 
 electro-magnet are connected directly to the dynamo- 
 terminals, through a fifty-candle power lamp and fuse 
 box, switches are not introduced, the circuits remaining 
 complete through the correctors, except in case of breakage. 
 The fifty-caudle power lamps, placed in the dynamo-room, 
 in addition to limiting the current through the magnet-coils, 
 by going out, will indicate a break in the circuit. A pre- 
 liminary experiment is required to ascertain approximately, 
 the number of turns of wire required in the winding, this 
 being known, the final adjustment is made, by moving them 
 as necessary in the slots in the securing brackets, until it is 
 ascertained by actual observation, that the compass remains 
 unafiected by the dynamo ; the correctors are secured in those 
 positions, and permanent connection made at the dynamo 
 terminals. 
 
 2. (Apollo Class.) Here two separate disturbing 
 elements exist, acting against, or in conjunction with one 
 another, according to whether the poles of the dynamo are 
 symmetrical to the middle line or not, with both dynamos 
 running, in one case the resulting error will be practically 
 nil, and in the other mil be the sum of the Separate errors. 
 
 The shunt coils of the. dynamo winding are the principal 
 exciting factors, by passing "a- current through the shunt 
 coils of a still (non-running -dynamo) from a working one it 
 can be excited, therefore, by arranging the shunt coils of the 
 two dynamos, so that a current can be passed through both, 
 from either working dynamo, both will be equally excited 
 and no disturbing influence will be exerted on the compass 
 if the necessary precautions are taken to connect up the 
 shunts, so that the poles of the dynamo are symmetrical to 
 the middle line. 
 
 Electric Lighting of Compass. — Owing to the marked 
 desire of many navigating officers to depart from the service 
 fitting and develop patents of their own, it may be as well 
 to briefly touch on systems tried but abandoned. The 
 original method naturally was to replace the oil lamp by an 
 
28 TEE MARINERS COMPASS. 
 
 electric one, the latter was fitted witli a flexible cable, from 
 a special kind of switch, with two terminals fixed to the 
 binnacle ; this was followed by the inside of the binnacle 
 being white enamelled, the card receiving the reflected light 
 from two 16 -candle power lamjps, placed under the compass 
 bowl, a lead-cased cable was carried up to each lamp and 
 separate switches were fitted on the binnacle ; and the third 
 was the intricate gibbet arrangement, jointed to freely turn 
 on its own axis, at the end of the arm was a metal case with 
 a hole in it, enclosing and of the same shape as a 16-candle 
 power lamp, through the hole in the metal case, a beam of 
 light was projected on to the lubber's line or any portion of 
 the compass card, as the arm was capable of being moved in 
 azimuth. 
 
 The present method is practically an improvement of the 
 first, and embodying the results of previous experience. In 
 the position usually occupied by the oil lamp, is placed an 
 electric 16-candle power one, enclosed in a metal cylinder, 
 with a spout* or tube, through which a beam of light passes, 
 sufficient to illuminate the lubber's point and the desired 
 portion of the graduation of the compass card. The current 
 is conveyed to the lamp by a twin cable, protected by 
 phosphor-bronze braiding,f consisting of two conductors of 
 twenty-three wires each of No. 38 L. S. W. G. The lamp 
 with flexible cable and plug piece of the portable connection 
 can be removed when not required ; the fixed part of the 
 connection is made watertight by means of a screwed metal 
 case. 
 
 The writer proposed in 1897 a method of compass 
 lighting and taking a bearing for small ships or 
 " Destroyers." Briefly, it consisted of two small electric 
 lights (half-candle power) similar to those in use for night 
 
 * The writer proposed a portable continuation of the tube, to prevent 
 the diffusion of the light before it reaches the upper surface of the glass 
 verge. 
 
 t Steel-braided cable is unsuitable on account of its magnetic properties. 
 
ELECTRIC LIQETING OF COMPASSES. 29 
 
 sights of guns, surmounting a small cylinder, containing an 
 8-candle power electric lamp (16-candle power was intended 
 for practical use), througli a narrow slit at each end of 
 cylinder, passed a beam of light sufficient to illuminate 10° 
 of arc of the card, the unit of bearing and (180° from it) 
 being denoted by a dark line : the line joining the electric 
 lights, the cylinder and the centre of the illuminated portions 
 of the card being in the same vertical plane. The 
 instrument working on a foot-plate similar to the one on 
 which the azimuth mirror works ; the elaborate fitting and 
 initial expense are against its introduction, but the writer 
 hopes it may form the groundwork for a future system of 
 lighting for that class of vessel. 
 
 The following caution should be carefully attended to, as 
 it is an important one, in fitting the cables and wires to a 
 compass pedestal, all conductors should be in pairs, side by 
 side, close together ; one in each pair, carrying the current 
 direct and the other the return ; so that equal currents 
 proceeding in opposite directions, will counteract each other 
 so far as their action on the compass needle is concerned. 
 
30 TEE MASmsH'S COMPASS. 
 
 CHAPTEE IV. 
 
 Deviation is the angle included between the magnetic 
 meridian and a vertical plane passing through the poles of 
 the compass needle, or in other words, the diflference between 
 the bearings shown by a shore compass (free from local at- 
 traction) and one on board ship. Deviation is caused by the 
 disturbing element of iron used in the ship's construction or 
 containeid within the ship, varies in different ships and in 
 various positions of the same ship ; in addition, the amount 
 varies at each compass, as the ship's head passes through the 
 different points of the compass. Deviation is ascertained by 
 " swinging," that is, causing the ship's head to complete 
 360° in azimuth, observing the bearing of a terrestrial or 
 celestial object from the standard compass, as the ship's head 
 is slowly placed on each compass point, comparing other 
 compasses with the standard by a bugle note, capable of 
 being heard throughout the ship, or by voice tubes. 
 
 The various methods are, first, by a well-defined object, 
 six to eight miles off (according to the size * of the diameter 
 of the circle made by the ship whilst being swung), of which 
 the magnetic bearing is known, and this can be obtained by 
 landing^ a compass on shore, the selected spot being free 
 from local attraction ; and observing the magnetic bearing of 
 the object when in line with the ship. The true bearing can 
 generally be ascertained from an Admiralty 'chart, if recently 
 surveyed, by applying the variation fOr the current year to 
 the true bearing, the magnetic bearing is obtained ; or the 
 
 * The position of the compass in the ship, whether near bow or stern, 
 especially in a long ship will materially affect the diameter of the circle, 
 
DEVIATION BY SWINQING. 31 
 
 horizontal angle measured from the well-defined object to 
 the sun, and its true bearing thus deduced.* With a ship 
 lying any length of time in a harbour, wind and tide com- 
 bined, will generally cause her to complete an entire circle in 
 azimuth, but when time will not permit of this being done, 
 the ship, is usually towed round by a tug and rope hawser. 
 At sea, a ship can be steamed round in a circle, observing 
 the bearing of the sun or a star ; two points in line (such as 
 leading marks), lighthouses in line, may be utilised for this 
 purpose ; the magnetic bearing is generally known and 
 shown on chart. 
 
 The difference between the magnetic and compass bearing 
 will be the deviation, and is called easterly or westerly 
 according to whether the compass north is drawn to the east 
 or west of magnetic north. 
 
 Second method — by Ke'ciprooal Bearings — is generally 
 used in a small confined harbour or in thick and misty 
 weather. In the former case care must be taken to ascertain 
 that the ship, whilst being swung, will not be affected by 
 the magnetic influencB of adjacent ships. A compass, pre- 
 viously compared with the standard, is landed in a suitable 
 spot, free from iron above or underground, or magnetic 
 rocks. To prevent possible errors, the shore compass and 
 standard should be clearly intervisible, and as an additional 
 precaution, watches should be compared. As the ship swings 
 round, she should be steadied on each point and at a pre- 
 concerted signal, such as the sudden dipping of a masthead 
 flag, the time noted, and simultaneous bearings taken of the 
 shore compass from standard, and of standard from shore 
 compass. Shortly before the shore compass is likely to be 
 hidden from view by masts, funnel or boats from the 
 standard, a horizontal angle should be measured (from the 
 
 * Azimuth tables for true bearing, by the late StaflF Commander J. 
 Burwood, E.N., extended by the late Captain J. E. Davis, or the Graphic 
 method of Captain Weir are published by J. D. Potter, 31, Poultry, 
 London. , A Variation Chart of the World is published by the 
 Admiralty. 
 
32 TEE MARINER'S COMPASS. 
 
 latter spot) between the shore compass and some well-defined 
 object, to prevent an important point being missed. If 
 only one compass is available, before swinging, land that 
 compass obtaining the magnetic bearing of a well-defined 
 object (nearly on a level with the standard is desirable), and 
 at each signal measure the horizontal angle between the 
 standard compass and the previously well-defined object ; 
 this wUl give the magnetic bearing of the former at each 
 observation. The difference between the shore bearing and 
 the ship bearing reversed (180° added) will be the deviation. 
 
 In fine weather at sea, in tropical countries, stellar 
 observations are generally employed for ascertaining the 
 deviation.* A circumpolar star at its greatest eastern or 
 western elongation, affords a ready method, as the star is 
 practically stationary in azimuth.f 
 
 The deviations are then entered in detail on a form 
 provided for that purpose. 
 
 The following points should be attended to, by whatever 
 method the ship is swung ; she should not be swung too 
 rapidly, but steadied on each point, because in the con- 
 struction of a ship, there is an intermediate class of iron, 
 neither soft nor hard, but which parts with its induced 
 magnetism slowly, and should an error be suspected from 
 this cause, the ship should be swung twice, but in opposite 
 directions, and the mean of deviations be used. 
 
 In addition, guns near the compass should be stowed for 
 
 sea, boom boats in their place, boat's davits turned in, and 
 
 cat davits down. In conning towers, where the top is 
 
 capable of being raised and lowered, its position should be 
 
 noted, whether the shutters are open or closed durijQg the 
 
 * A difficulty may occasionally occur in manipulating the electric light 
 for stellar observations. 
 
 I Sin. az. = sec. lat. x cos. dec. 
 
 Sin. alt. = sin. lat. X cosec. dec. 
 Cos. hour angle = tan. lat. X cot. dec. 
 
 A star can easily be identified if the above simple solutions in right 
 angle spherical trigonometry be previously worljed out. 
 
TABLE OF DEVIATIONS. 
 
 33 
 
 swinging, also when the conning-tower compass is being 
 compared with the standard. Colliers or water-tank vessels 
 should not be alongside during swinging. 
 
 If a point be missed, it is usually interpolated, but if the 
 errors are large they can be plotted on prepared paper ruled 
 in small squares, in a curve, or on Napier's diagram ; the 
 latter is a straight line, divided into 32 or 360 equal parts, 
 representing the points or degrees of the card ; the devia- 
 tions are laid off on lines, cutting the vertical one at an 
 angle of 60°, easterly deviations on the right, and westerly 
 deviations to the left of the vertical line. On the Table of 
 the deviations should be placed the position, size, and 
 number of the fore and aft, athwart and vertical magnets. 
 
 Below is a specimen of a Table of deviations obtained by 
 a ship swinging from a buoy with a distant object, and one 
 underweigh with the sun : 
 
 H.M.S. "Waespite," Sheerness. 
 
 H.M.S. " AeiifCOUKT," Nore. 
 
 Magnetic bearing of Jezreel Tower, 
 
 Apparent time of place has been 
 
 S. 78° 20' W., distant 8 miles. 
 
 omitted. 
 
 Direction 
 of Ship's 
 Head by 
 
 Standard. 
 
 Bearing by 
 
 Compass of 
 
 Jezreel Tower. 
 
 Deviation. 
 
 Direction 
 of Sliip's 
 Head by 
 Standard. 
 
 Magnetic f 
 
 ■ Bearing of 
 
 the Sun. 
 
 Bearing by 
 
 Compass of 
 
 the Sun. 
 
 Deviation. 
 
 
 o / 
 
 O 1 
 
 
 / 
 
 o t 
 
 f 
 
 North 
 
 S. 78 00 W. 
 
 20 E. 
 
 North 
 
 S. 68 30 W. 
 
 S. 68 SOW. 
 
 00 
 
 N.N.E. 
 
 S. 79 50 W. 
 
 1 SOW. 
 
 N.N.B. 
 
 S. 77 00 W. 
 
 S. 76 00 W. 
 
 1 00 w. 
 
 N.E. 
 
 S. 68 50 E.* 
 
 1 10 W. 
 
 N.E. 
 
 S. 76 30 W. 
 
 S. 76 00 W. 
 
 sow. 
 
 E.N.E. 
 
 S. 69 00 E.* 
 
 1 00 W. 
 
 E.N.E. 
 
 S. 76 00 W. 
 
 S. 75 SOW. 
 
 sow. 
 
 East 
 
 8. 69 00 E.* 
 
 1 00 W. 
 
 East 
 
 S. 75 30 W. 
 
 S. 75 00 W. 
 
 sow. 
 
 E.8.E. 
 
 S. 77 00 W. 
 
 1 20 E. 
 
 E.S.E. 
 
 S. 74 00 W. 
 
 S. 75 00 W. 
 
 1 00 E. 
 
 S.E. 
 
 S. 77 45 W. 
 
 45 E. 
 
 S.E. 
 
 S. 72 SOW. 
 
 S. 74 00 W. 
 
 1 30 E. 
 
 S.S.E. 
 
 S. 78 00 W. 
 
 20 E. 
 
 S.S.E. 
 
 S. 71 SOW. 
 
 S. 73 80W.'2 00 E. 
 
 South 
 
 S. 77 45 W. 
 
 45 E. 
 
 South 
 
 S. 71 00 W. 
 
 S. 73 00 W. 
 
 2 00 E. 
 
 s.s.w. 
 
 S. 77 30 W. 
 
 50 E. 
 
 S.S.W. 
 
 S. 70 40 W. 
 
 S. 72 0(1 W. 
 
 1 20 E. 
 
 s.w. 
 
 S. 78 50 W. 
 
 SOW. 
 
 s.w. 
 
 S. 70 40 W 
 
 S. 71 SOW.'O 50 E. 
 
 w.s.w. 
 
 S. 80 00 W. 
 
 1 40 W. 
 
 w.s.w. 
 
 S. 70 20 W. 
 
 S. 71 OOW.U 40 E. 
 
 West 
 
 S. 80 00 W. 
 
 1 40 W. 
 
 West 
 
 S. 69 40 W. 
 
 S. 70 40 W. 1 00 E. 
 
 W.N.W. 
 
 S. 79 45 W. 
 
 1 25 W. 
 
 W.N.W. 
 
 S. 69 00 W. 
 
 S. 70 OOW.'l 00 E. 
 
 N.W. 
 
 S. 79 00 W. 
 
 40 W. 
 
 N.W. 
 
 S. 68 00 W. 
 
 S. 69 SOW.,1 30 E. 
 
 N.N.W. 
 
 S. 78 00 W. 
 
 20 W. 
 
 N.N.W.S. 67 30 W. 
 
 S. 68 45W.1 15 E. 
 
 * Jezreel Tower obscured, Minster Church observed. 
 
 t It is usual in practice, before swinging, to convert the sun's true into magnetic 
 bearing for the period of time, which will probably be occupied in adjusting and 
 swinging the ship. 
 
34 THE MABINEB'S COMPASS. 
 
 In addition to the above methods of ascertaining the 
 deviation by bearings of visible objects, several instruments 
 have been devised for correcting the error of a compass, 
 without bearings, an invaluable assistance in thick weather. 
 The inventions of Sir William Thomson (Lord Kelvin),* 
 Dr. Waghorn deflectors,* and Chevalier Peichl Control com- 
 pass,! ^^^ fo^ t^8'* purpose. If the directive forces acting on 
 the compass were equalised, there would be no deviation ; 
 but if an instrument were devised capable of measuring the 
 directive force of the compass needle with the ship's head on 
 the magnetic cardinal points, their comparative strength 
 would be known and could be equalised by fore and aft and 
 athwartship magBtets. The horizontal force acting on the 
 compass, due to the hard, iron in the ship, will either increase 
 or lessen the normal amount, according to whether it is 
 acting with or against the horizontal component of the 
 earth's magnetic force ; in the former case the angle of 
 deflection produced by another magnet of a certain power, 
 at a certain position and distance on the compass needle, 
 will be smaller than in the latter case, because the smaller 
 the directive power of the needle, the larger the deflection. 
 Now, supposing with a certain magnetic force we always 
 deflect the compass needle 90° on the four cardinal compass 
 points, it would be found that increased or decreased mag- 
 netic power would be required, for if only the normal force 
 were exercised on all points, the directive force would be 
 equal, and there would be no deviation. A deflector 
 measures the augmented or diminished power, and enables 
 a mean to be taken, at which the deflector is set, the 90° 
 deflection being obtained by moving the fore and aft and 
 athwartship magnets. In the above explanation, the quad- 
 lantal error | is supposed to have already been corrected, 
 
 * Fully described in ' Deviation of the Compass,' by Professor Eeinold & 
 Dr. Waghorn. Published by Henry Eichardson, 4, Church Street, Greenwich. 
 
 t Described by Staff-Commander (now Captain) Creak in No. CXVII. 
 of the Journal of the Eoyal United Service Institution. 
 
 I For Quadrantal error, see page 41. 
 
mSTVBBJNG EFFECT OF IRON. 35 
 
 and as after being corrected it remains constant for all 
 latitudes, it is unlikely that the deflector would be used 
 for the purpose of correcting quadrantal error,- but the 
 normal deflection (90°) can be obtained by moving the 
 quadrantal spheres. 
 
 Chevalier Peichl Control compass is based on the fact 
 that when a dip 6ircle is placed in the magnetic meridian, 
 the minimum dip is obtained. The principle only on which 
 the above deflectors act is briefly stated; it requires an 
 expert practical observer to use them with advantage, but 
 in his hands they have given satisfactory results. 
 
 We will now proceed to inquire into the eff"ect of the 
 iron in a ship on the compass, the general conclusions to be 
 arrived at, and any known disturbing elements, whether 
 regular or irregular. In the early wood-built ships, with a 
 compass in the centre line of the ship, placed on the quarter- 
 deck, the points of no deviation were with the ship's head 
 near north and south, and the maximum deviation when 
 east and west, easterly deviation with the ship's head in the 
 eastern semi-circle, and westerly when in the western semi- 
 circle, but becoming reversed in southern magnetic latitudes. 
 In wood-built steam-vessels the same general features 
 occurred, except the points of maximjim and minimum 
 deviations were near, but not at;, the cardinal points, the 
 deviation usually becoming , reversed in southern magnetic 
 latitudes. In both ,cases.a Flinders bar could have been 
 employed, if considered necessary. In the composite, and 
 early iron-built ships, the points of nq deviation, instead of 
 being the same as in wood-built steam-vessels, were chiefly 
 dependent on the magnetic directipn in which the ship was 
 built-;-the ship becoming a magnet, or, strictly speaking, 
 the various parts and fittings of the ship becoming a series 
 of small magnets with their red and blue poles parallel, the 
 magnetism being more or less permanent, according to the 
 degree of, hardness of the iron used and the amount of 
 mechanical violence to which it is subjected, but it is more 
 
 P 2 
 
36 
 
 THE MARINER'S COMPASS. 
 
 convenient to regard the ship as a whole when dealing with 
 permanent magnetism. The ship, whilst being built, becomes 
 a magnet from the induced magnetism of the earth ; the 
 greatest force is the line of the earth's dip at the place, 
 being also the points of maximum and , minimum directive 
 force of the compass needle ; there is a neutral line at right 
 angles to the dip. Below is shown the distribution of 
 magnetism for an iron ship built in England on the cardinal 
 points (Figs. 1, 2, 3, 4). 
 
 The disturbance caused by permanent* magnetism on 
 the compass needle can be clearly illustrated by placing a 
 magnet exercising less power than the horizontal component 
 
 F/j J. Built Head East 
 
 Fit^.f-. Built Head West. 
 
 of the earth's force near a small compass, and causing the 
 magnet to complete a revolution around the compass ; the 
 disturbances produced will depend on the relative values of 
 earth and magnet. With the Standard compass placed well 
 forward, and the introduction of iron turrets, conning-towers, 
 masts, chart-houses, superstructures, and in the ship's fittings 
 generally, the direction in which tie ship was built is no 
 longer clearly discernible, but is lost in the disturbances 
 produced by the surrounding iron. It is customary, where 
 
 * In order to prevent confusion in the mind of the reader between 
 "permanent" and "sub-permanent" magnetism, the former is generally 
 used in the sense of being the opposite of transient magnetism. .The 
 expression sub-permanent magnetism was introduced by the late Sir G. B. 
 Airy to^ show the lesser degree of permanency in the so-called permanent 
 magnetism of an iron ship to that possessed by a magnet ; both change, 
 but in the latter case very slowly. 
 
SUB-PEBMANENT OB UNSTABLE MAGNETISM. 37 
 
 practicable, to place the armour-plates in position, and 
 complete the ship in the reserve with the ship's head in an 
 opposite direction to her building one, with a view to reduce 
 her semi-circular deviation,* the magnetic direction of the ship 
 will then be intermediate between the direction of the ship's 
 head whilst being built, and that whilst being completed for 
 commission. It should be borne in mind the greatest change 
 in the magnetic character of a ship occurs after launching 
 until she is commissioned, but now that ships are 
 subjected to force, natural, and consumption steam trials, 
 in addition to gun trials, it may be fairly assumed that, 
 on commissioning, the magnetic character of the ship has 
 become fairly stable, f 
 
 Sub-Permanent J magnetism produces deviation in a 
 similar manner to the permanent, but it may be increased 
 or diminished, or even reversed, in sign. In large ships, 
 with a Standard compass properly placed, the change is 
 small, but in " Torpedo-boats " and " Destroyers " it has 
 varied 6° in a comparatively short space of time. It is due 
 to that intermediate class of iron neither hard nor soft, but 
 possessing some of the characteristics of both, and is liable 
 to change owing to lapse of time, or suddenly reversed by 
 severe shocks, such as gun-firing (the disturbance caused by 
 heavy gun-firing might be diminished by firing the guns with 
 the ship's head in opposite directions), the error from this 
 cause may disappear in an hour, or the return to the normal 
 deviation may be gradual, occupying some days. An inter- 
 mediate class of soft iron may be temporary magnetised by 
 induction through the vibration of a ship when steaming at 
 high speeds ; steaming at ordinary speed against a head sea ; 
 docking ; steaming, or remaining on a given course for a 
 
 * Semi-circular deviation, see page 40. 
 
 t In the mercantile marine the " stable period " is generally arrived at ' 
 within the year, and nearly so after a few months' active service. 
 
 X See note * on previous page. Sub-permanent is used in these pages 
 to denote an Unstable period. Unstable magnetism would be a better term 
 than Sub-permanent. 
 
38 TEE MARINERS COMPASS. 
 
 considerable time these disturbances are usually called Reten- 
 tive (more correctly Transient) error and Gaussin error. 
 
 Gaussin error should always induce great caution and 
 watchfulness, because, if after several days' steaming on one 
 course, the course is suddenly altered to one at right angles 
 to the former one, perhaps in thick weather, the danger is 
 that the disturbance does not begin to take effect on the 
 needle until the second course is being steered, although it 
 is gathering in strength whilst on the original course ; and 
 again, the disturbance gradually diminishes whilst on the 
 second course, but not according to any known law. No 
 corrector, at present, exists for this error, but a navigating 
 ofEcer, by carefully watching the effects produced by the 
 above causes in fine weather, will be in a position to predict 
 in a general manner the probable disturbance in thick 
 weather, and this will give the navigating officer greater 
 confidence.* The disturbance is greater in high magnetic 
 latitudes than in low, and when the original course is east 
 and west than when north and south, the tendency being for 
 the ship's head to be drawn in the direction of the original 
 course. In preparing a Deviation table by swinging, Gaussin 
 error can be eliminated by swinging first to starboard and 
 secondly to'port.f The following is taken from the instruc- 
 tions for Sir William Thomson's compass:* "If the ship 
 underweigh is steamed round on the different courses, the 
 amount of the ' Gaussin error ' may generally be greater 
 than if she is hauled round by warps ; but we must not be 
 sure that it will be so, because the shake of the screw, which 
 enhances the magnetisation on the east or west courses, may 
 shake it out again before the observation is made on the 
 north and south courses. "J 
 
 * . See instructions for Sir William Thomson's, (Lord Kelvin) compass. 
 James White, 209, Sauchiehall Street, Glasgow. 
 
 t Captain Mayes states that the " Defence " swung ia.l865 in this manner, 
 the maximum difference was 3° on the north and south points. 
 
 t A (laussin error of 3° developed by second-class cruisers whilst lying 
 alongside the basin wall at Chatham during the period they were in the 
 
. TBAN8IENT MAGNETISM. 39 
 
 Torpedo-boats and Destroyers are supplied with the 
 improved liquid compass as a Standard, placed in a selected 
 position, but only five feet* from a thin iron or steel struc- 
 ture, subject to excessive vibration, and an occasional con- 
 cussion from firing of guns. It is, therefore, to be expected 
 that their magnetism, whether permanent or sub-perma- 
 nent, will vary considerably more than in larger ships. In 
 Torpedo-boats, the close agreement of the deviations, after 
 considerable time has elapsed, is remarkable, but it is pos- 
 sible that their magnetic character at sea may alter, reverting 
 to the normal (harbour) character after being in harbour 
 some days. In some of the early "Destroyers," a rapid 
 change in the deviation took place, owing to an equally 
 rapid change in the sub-permanent magnetism of the ship ; 
 but with the increased height of the standard compass from 
 the deck, this error has been diminished, but constant 
 watchfulness and observation should be exercised after guns 
 are fired, and the disturbances described previously under 
 " Transient error " apply with increased force to " Destroyers " 
 and " Torpedo-boats," but if a careful sea magnetic record were 
 kept in " Destroyers," under the varying circumstances 
 described above, it is probable some practical rules could 
 be deduced therefrom, and issued for the benefit of their 
 navigating officers. 
 
 There is an irregular disturbing element to the compass 
 outside the ship, viz., the land itself, not above the surface 
 of the water, but below it ; for many years past, it has 
 been reported that the compass was disturbed on passing 
 islands, points, etc. ; but the reports were discredited, as 
 owing to the laws of magnetic attraction, even if the nearest 
 visible land were composed of magnetic ore, it could not 
 exercise any appreciable disturbing eff'ect on the compass at 
 the distance it was being passed. The Surveying schooner 
 
 Fleet Eeserve, generally disappeared on arrival at Sheerness under stoam, a 
 distance of 8^ miles. 
 
 * In Destroyers this distanca is now increased, to 6 feet. 
 
40 THE MARINER'S COMPASS. 
 
 Meda* in 1883-85, by careful observations off tbe island 
 of Bezout — north-west coast of Australia, the land being 
 distant about three miles — showed that the source of the 
 disturbance was submarine, viz., that the bottom travelled 
 over (although in eight fathoms water), was sufficiently- 
 magnetic for a mile to disturb the needle 30° f Similar 
 disturbances have been reported from Labrador, Mada- 
 gascar, Solomon Islands, Iceland, Java, Odessa Bay, and 
 Isle de Los, West Coast of Africa ; an extra caution should 
 therefore be exercised by the jiavigator when in those 
 localities. 
 
 Semi-Circdlar Error is so called from being easterly 
 ( -f ) in one semi-circle, and westerly ( — ) in the other. The 
 larger portion is caused by the permanent magnetism of the 
 ship (the small changing portion in the sub-permanent 
 magnetism may be expected to diminish very slowly with 
 lapse of time, and for which no corrector has been found). 
 As previously mentioned, there will be points of no devia- 
 tion, or neutral points, 180° apart, the deviation for any 
 point will be the maximum deviation multiplied by the sine 
 of the azimuth, from the neutral point of the disturbed 
 compass. On change of geographical position, this portion 
 will vary inversely as the earth's horizontal force, the per- 
 manent magnetism exercising least force at the magnetic 
 equator, and again increasing, according to the rule given 
 above, as the ship proceeds south, or increases her magnetic 
 latitude. This force may be considered as acting in the ship 
 in a fore and aft and athwartship direction, and the errors 
 produced are therefore corrected by permanent magnets 
 placed in those positions in the binnacle. The remaining 
 portion of the semi-circular deviation is caused by induced 
 magnetism in vertical soft iron. In England (northern 
 
 * Commanded by Staff-Commauder (now Captain) J. E. Coghlan. 
 
 t Captain W. Usborne Moore, in H.M.S. Penguin, re-examined this 
 locality, obtaining a maximum disturbance of 55°, the disturbed area 
 remaining about the same. 
 
SEMICIBGULAB EBBOB. 41 
 
 magnetic latitudes), the lower ends possessing red, and the 
 upper blue magnetism, varying as the tangent of the dip,* 
 disappearing at the magnetic equator ; but as the ship pro- 
 ceeds from there, south, blue magnetism is developed in the 
 lower ends, and red in the upper, increasing in strength 
 according to the tangent of the dip. With a compass in the 
 centre line of the ship, the iron is fairly symmetrically dis- 
 tributed with the fore and aft line, and the error is therefore 
 corrected by vertical soft iron (Flinders Bar) placed in the 
 centre, of the fore or aft side of the compass as necessary. 
 As a general rule with compasses placed well forward or aft ; 
 the disturbing influence of vertical iron is abaft or forward 
 of compasses so placed ; the Flinders bar is therefore placed 
 on the opposite side of the compass to which the disturbing 
 influence is exerted, in order to counteract it. Cases may 
 occur where the disturbing influence of the ship's vertical 
 iron, on the compass, may be dominated or controlled by 
 the proximity of a conning tower, funnels, iron masts, etc. 
 From the foregoing remarks, it will be readily seen that, if 
 a ship, after rough adjustment, proceeds to a place with 
 greater horizontal force, increased power in fore and aft 
 magnets will be required; if to greater vertical force, in- 
 creased Flinders bar will be required, f 
 
 QUADRANTAL ERROR is due to induction in horizontal soft 
 iron,J continuous or non-continuous, fore and aft and athwart- 
 ships, and is so called from its being alternately easterly and 
 
 * See page 18 for connection between horizontal, vertical, total force 
 and dip. 
 
 t Lord Kelvin's (Sir William Thomson) Flinders bar is about 24 inches 
 long, composed of cylindrical iron in separate pieces of unequal length, the 
 longest being on top, which is kept at a distance of 2 inches above and 
 about 9 inches from the centre of the compass. To keep the vertical 
 height constant, duplicate pieces are made of wood and placed at the 
 bottom of the brass tube containing the Flinders bar. 
 
 J Non-continuous usually reckoned -|-, and continuous — . A ship 
 whoUy armour-plated will tend to diminish the quadrantal error and 
 strongly diminish the directive force. A ship with unarmoured ends, but 
 with armour bulkheads near the compass, will produce a large quadrantal 
 error, but only slightly diminish the directive force. (S3e Fig. 8.) 
 
42 THE MABINERS COMPASS. 
 
 westerly in the four quadrants ; its eflfect can be practically 
 shown by carrying a bar of soft iron in the same horizontal 
 plane with and around a small compass ; at the cardinal 
 points it will exercise no disturbing influence on the needle , 
 and the maximum points of disturbance will be at the inter- 
 cardinal points. In swinging a ship from north through east, 
 the deviation is usually easterly ( + ) in the N.E. and S.W. 
 quadrants and westerly ( — ) in the S.E. and N.W. quadrants, 
 i.e. + , — , + , — . Continuous masses athwartships and non- 
 continuous masses in the fore and aft line cause an easterly 
 (4-) deviation in the N.E. quadrant, and continuous masses 
 in the fore and aft line and non-continuous athwartships 
 produce a westerly ( — ) one in the same quadrant — the 
 iron beams used in the ship's construction being the principal 
 source of quadrantal error, it is therefore connected by 
 detached soft iron masses athwartships. The usual cor- 
 rectors are hollow* nouTmagnetic cast-iron spheres, of a 
 minimum thickness of an inch, mounted on brass brackets, 
 the spheres are equidistant from, and their centres in the 
 same horizontal plane with the centre of the needles, f 
 With a compass, not in the midship line ; or a compass in 
 the midship line, with iron unsymmetrically distributed, such 
 as turrets en echelon J ; the spheres being still equidistant and 
 in the same horizontal plane as the compass needles, must 
 be placed at an angle with the athwartship line. In case, 
 when swinging from north through east, a — (negative) § 
 quadrantal error is found, i.e., — , 4-, — , +, it may arise 
 from over-correcting a previous -|- (positive) error ; or with a 
 mast or pole compass, when it is placed near the centre of, 
 the ship's length ; but the amount in each case will be small. 
 
 * Hollow spheres are employed on account of their lightness ; their 
 value in correcting quadrantal error is for practicable purposes equal to 
 solid spheres, provided the thickness of their shell exceeds an inch. 
 
 t In using the expressions vertical and horizontal refers to when the 
 ship is on an even beam, i.e., upright, unless the contrary is expressly stated. 
 
 t H.M.S. Inflexible. 
 
 § — P was often found in shallow wood-built screw ships. 
 
QUADS ANTAL AND SEELING ERRORS. 43 
 
 QuADRANTAL ERROR alters slightly by lapse of time, but 
 will not vary on change of magnetic latitude ; the increased 
 or decreased strength of the induced magnetism in horizontal 
 soft iron (which is the cause of the quadrantal error) will 
 vary in the same proportion to the increased, or decreased 
 horizontal force directing the needle, and the resultant of 
 the forces will not vary in direction ; these remarks do not 
 apply to. the Peichl compass- corrector, its soft iron rod 
 correctors acquire temporary induced magnetism from the 
 needles and are altered on change of magnetic latitude. 
 Quadrantal spheres increase the needles' directive force on 
 east and west courses. 
 
 Heeling Error. — Hitherto the deviation has been of a 
 ship on even beam, that is upright, and the vertical com- 
 ponent of the ship's permanent magnetism caused no 
 deviation. If a ship, roughly connected for quadrantal 
 error, is heeled, the vertical force being no longer so, will 
 cause an error (1) acting upwards or downwards,* according 
 to the direction in which the ship was built, the attraction 
 of the needle to the windward or leeward side depending on 
 the position of the compass in the ship. (2) Horizontal 
 soft iron, consisting of beams, decks, etc., and horizontal 
 detached masses become inclined, receiving temporary 
 induced magnetism from the earth's vertical force ; their 
 upper ends with blue and lower ends with red magnetism in 
 northern, and upper ends red and lower blue in southern 
 magnetic latitudes.! The continuous and detached hori- 
 zontal soft iron will act in an opposite direction, and in 
 addition the quadrantal spheres will tend to compensate a 
 portion of the heeling error due to continuous horizontal soft 
 iron. The above are the principal causes of the heeling 
 error, and it will be readily seen that they exercise their 
 greatest force on the needle when the ship's head is compass 
 
 * Usually marked - if acting upwards, and -|- if downwards, 
 t Changing their signs on crossing the magnetic equator and increasing 
 in power as higher magnetic latitudes are reached. 
 
4-i THE MARINER'S COMPASS. 
 
 north and south, being then at right angles, and their 
 minimum when parallel, i.e., with the ship's head compass 
 cast and west. The Heeling error is usually counteracted 
 by magnets placed end on, vertically below the centre of the 
 compass ; it is better to employ two or three small magnets 
 instead of a large one, increasing their number if necessary 
 in lieu of decreasing the distance between the magnets and 
 the compass.* Vertical induction in vertical soft iron (iron 
 stanchions, funnels, masts) the direction and intensity will 
 vary when the ship heels, this change is generally merged 
 into that due to the ship's vertical force acquired whilst 
 building, and they are treated as one. Vertical masses 
 above the compass act in a contrary manner to similar 
 ones below the compass, thus tending to diminish the 
 heeling error or the proximity of a mast or funnel to 
 the compass, may counteract and exceed the combined 
 disturbances of the vertical force of the ship and vertical 
 soft iron. This error can be corrected by a soft iron bar, or 
 by raising the upper end of the Flinders bar above the 
 level of the compass. 
 
 A small heeling error not usually corrected is caused 
 by the. (soft iron) of the keel, inducing a transient vertical 
 force, which has no disturbing influence when the ship's 
 head is on the east and west points ; it is therefore desir- 
 able to correct the heeling error when on one of those 
 points. Practically what is done in correcting the heeling 
 error is to make the vertical and horizontal forces at the 
 compass bear the same ratio as the vertical and horizontal 
 forces do on shore. It should also be borne in mind that 
 the healing error is not wholly corrected, but the error is 
 brought within reasonable limits. 
 
 Heeling Co-efficient is the amount of deviation caused 
 
 by 1° of heel when the ship's head is north or south by the 
 
 ship's compass, the amount then being at its maximum ; and 
 
 is found for any other azimuth by multiplying it by the 
 
 * Alters on change of magnetic latitude, i.e., of vertical force. 
 
CONSTANT DEVIATION. 45 
 
 cosine of the azimuth measured from the points of maximum 
 disturbance. 
 
 Constant deviation is the amount the easterly ( + ) 
 deviations exceed the westerly ( — ) or vice versd. In a 
 properly placed compass it is generally under 1°, but in a 
 compass out of the centre line of the ship and near un- 
 symmetrically placed soft iron, it will exceed that amount. 
 It does not change with the magnetic latitude and remains 
 constant on all courses, in a steering compass it can be 
 corrected by altering the lubber's line. A false or apparent 
 constant deviation may be due to instrumental errors in the 
 compass or azimuth mirror, or from an incorrect magnetic 
 bearing. 
 
46 TEE MAMINES'S COMPASS. 
 
 CHAPTER V. 
 
 In the previous Chapters, how to obtain the deviation and 
 the causes of it have been explained. We will now propbse 
 to analyse the deviation to show the constituent parts of the 
 erroK. All forces acting on the compass needle may be 
 resolved in horizontal and vertical planes passing through 
 the centre of the needles ; as already shown the vertical 
 forces wUl produce no errors so long as the ship remains 
 upright ; the horizontal force can be divided into two forces, 
 one acting in a fore and aft, and the other in athwartship 
 direction (Fig. 5).* To analyse a deviation table, the late 
 Mr. Archibald Smith represented their constituent parts by 
 the letters A, B, C, D, and E, called co-efficients ; using the 
 following equation, if S be the deviation for any compass 
 course 6, reckoned from to 360, through east — 
 
 S = A + B. Sin ^ + C. Cos ^ + D. Sin 2 6* + E. Cos 2 9. 
 
 The co-efficients shown in degrees and minutes represent 
 the maximum error produced separately by the disturbing 
 forces of the ship's magnetism. If the errors are small, the 
 forces are considered proportional to the errors they produce, 
 but strictly, they are proportional to the sines of those 
 errors. 
 
 A is the constant! error (real) due to the horizontal 
 induction in soft iron unsymmetrically placed, and is ascer- 
 
 * In Fig. 5, E represents the combined disturbing horizontal force 
 of the ship's iron on the compass. It can be resolved into the forces 
 A (C) and P (B). 
 
 t For Constant error, see page 45, 
 
S AND a CO-EFFICIENTS. 47 
 
 tained by adding algebraically the deviations (easterly 
 deviations + , westerly deviations — ) on the four cardinal 
 points, and dividing their sum by four. 
 
 B and C represent the semi-circular * error ; B is called + 
 when the force is towards the bow, and — when towards the 
 stern ; C is called + when the force is towards the starboard 
 side, and — when towards the port side. B is approximately 
 the deviation at east ; or west with its sign changed, or the 
 mean of the two. C is approximately the deviation at north ; 
 or the deviation at south with its sign changed, or the mean 
 of the two. B and C represent, respectively, the magnetic 
 fore and aft, and athwartship forces of the ship. - 
 
 iL ^ 
 
 J R 
 
 Fio. 6. Fig- 5. 
 
 ■ In Fig. 6 (ship's head north), B will have no disturbing 
 effect, as it is acting in line or parallel to the direction of 
 the needle ; but the C force, being at right angles, will be at 
 its maximum. 
 
 In Fig. 7 (ship's head east), the force of B being at right 
 angles, will be at its maximum and C at its minimum, the 
 force being parallel to the needle. As the forces producing 
 semi-circular error may be considered porportionate to the 
 deviation that takes place. Semi-circular error is ex- 
 pressed by the second and third terms of the equation : 
 B Sin ^ -f C Cos e. 
 
 D and E are the component parts of the quadrantal 
 
 error, f D is the mean value of the deviations at N.E. and 
 
 S.W. ; or the deviations at S.E. and N.W., with their signs 
 
 cha,nged, or the mean of both means ; and is the maximum 
 
 * For Semi-Circular error, see page 40. 
 t For Quadrantal error, see page 41, 
 
48 
 
 TEE MARINER'S COMPASS. 
 
 quadrantal error caused by teraporary induction in horizontal 
 soft iron symmetrically placed. E is the maximum quad- 
 rantal error caused by temporary induction in horizontal 
 soft iron unsymmetrically placed, is closed allied to A (real) ; 
 and it is found by adding algebraically the deviations on the 
 four cardinal points, reversing the signs on east and west, 
 and dividing their sum by 4. 
 
 D Fis. 8. — Horizontal soft iron symmetri- 
 cally placed. Continuous masses in 
 tlie fore and aft line and non-con- 
 tinuous athwartsMps. Cjiuse — 
 (westerly) error in N.E. quadrant. 
 
 E Fig. 9. — Horizontal soft iron un- 
 Bjnnmetrically placed. Turrets en 
 Echelon. 
 
 Usually for a standard compass, E is very small in 
 amount and remains uncorrected, but in a ship with turrets 
 en echelon, and in protected stations between decks with the 
 compass out of the centre line of the ship, it may be 
 necessary to compensate for the error, by placing the line 
 joining the centre of the quadrantal spheres at an angle fi 
 
 E* 
 
 with the athwartship line, so that tan 2/8 = =^ 
 
 In a 
 
 similar manner in " Destroyers," where the poles of a dynamo 
 make an angle with the athwartship line, the soft iron 
 corrector in the deck is placed at a like angle. Quadrantal 
 error will be expressed by the 4th and 5th terms of the 
 equation : D Sin 2^ + E Cos 26,\ and— 
 
 8 = A + B. Sin ^ + C. Cos ^ + D. Sin 26 + E. Cos 2d. 
 
 Below is an example of obtaining the co-efficients % from 
 
 * ' Deviation of Compass,' by Professor Eeinold and Dr. Waghorn, 
 page 19. 
 
 t For mathematical proof, Hid., page 17. 
 
 X The method of obtaining the co-efficients from deviations observed 
 on 32 or 16 points of the compass is shown in the Admiralty Manual. 
 The general method is the same. 
 
CO-EFFICIENTS. 49 
 
 the deviation table of H.M.S. Warspite (already given at 
 page 33), from eight points, viz., on North, N.B., East, S.E., 
 South, S.W., vfest, and N.W. Mark six columns I, II, III, 
 IV, V, and VI ; place the deviations on the first four points 
 under I, and the remaining four under II, with their proper 
 signs prefixed. 
 
 III will be half the sum (algebraically) of lines under I 
 and II. 
 
 IV will be half the sum (algebraically) of lines 
 under I and II, but the signs in lines under II must 
 be changed. 
 
 Under V and VI, place the values under III after being 
 multiplied by the natural sines and cosines (0° and 45°). 
 
 Half the algebraic sum under V and VI are the co- 
 efficients B and C. 
 
 Mark four columns Ia, Ha, IIIa, and IVa, the two 
 upper lines under III place under Ia, and the two lower 
 under IIa. 
 
 Place half the algebraic sum under IIIa and again 
 under IVa, but in the latter case the signs under IIa must 
 be changed. 
 
 The upper line will be E and the lower D. 
 
 Half the algebraic sum under IIIa will be A. 
 
 I. II. III. IV. V. VI. 
 
 U/ O/ O/ 0/ "^ o/ 
 
 + 20 + 45 + 82 - 12 00 ' - 12 
 
 - 1 10 - 30 - 50 - 20 - 14 - 14 
 
 - 1 00 - 1 40 - 1 20 + 20 + 20 - 00 
 + 45 -0 40 +0 02 +0 42 + 29 (- +) - 29 
 
 + 49 2) - 55 
 
 - 14 - 27 C. 
 
 2) + 35 
 
 B is + 17 
 
50 TEE MARINER'S COMPASS. 
 
 lA. IlA. IIlA. IVa. 
 
 at Of O I O / 
 
 + 32 - 1 20 - 24 + 56 is E 
 
 - 50 + 02 - 24 - 26 is D 
 
 - 48 
 
 - 24 is A. 
 
 To obtain the deviation from the co-efficients is to make 
 use of the late Mr. A. Smith's formula at the commencement 
 of this chapter.* This following formula f will give B, C, D, 
 by observations made in one quadrant ; A and E are assumed 
 to be nil. The letters NE, N, and E, prefixed to 8, refer 
 to the deviation with the compass, respectively, on those 
 points. 
 
 N E auadrant D - Sii^ ^5° X Sin N E 8 - ^ (Sin N 8 + Sin E 8) 
 N.K quadrant. D _ gj^ 45° X Cos N B 8 - i (Sin N 8 - Sin E 8) 
 
 B = (l + D). 8inE8. = (1 - D). Sin N 8 
 
 o j; p. ^ - Sin 45° X Sin S E 8 + h (Sin E 8 + Sin S 8) 
 
 ■ ■ " Sin 45° X Cos S E 8 - i (Sin E 8 - Sin S 8) 
 
 B = (1 + D). Sin E 8. C = - (1 - D). Sin S 8 
 
 c, ^ ^ _ Sin 45° X Sin S W 8 - ^ (Sin S 8 + Sin W 8) 
 
 ■ ■ " Sin 45° xCosSW8-4(Sin 88- Sin W8) 
 
 B=-(l + D). SinW8. C = - (1 - D). SinS8 
 
 ^ ^ „ _ -Sin45°x SinNW8 + ^(SinW8 + Sin2Sr8) 
 
 " Sin45°xCosNW8-i(8inW8-SinN8) 
 
 B = (1 + D). Sin W 8. C = (1 - D). SinNS 
 
 The directive force of the needle in the earlier ironclads 
 was found to have been reduced about a fourth in amount ; 
 with the introduction of conning towers and compasses placed 
 near and inside them, the directive force wag still further 
 reduced^ until it was only one-fourth of the amount in com- 
 parison with the shore. In preAdous paragraphs referring to 
 permanent magnetism J and semi-circular { error, it was 
 shown that the effect of the permanent magnetism on the 
 
 *_ See Admiralty Manual, page 27. t ^^id, page 38. 
 
 X See pages 36 and 40 respectively. 
 
X (LAMBDA). 51 
 
 directive force of the needle is to augment its normal amount 
 on one point and to decrease it by an equal amount on the 
 opposite one \ 180° removed from one another, according to 
 whether it is acting with or against the earth's force ; the 
 mean value remaining unaltered ; it is, therefore, apparent 
 the diminution in the directive force of the needle on board 
 ship must be due to some other cause, viz., to horizontal in- 
 duction in soft iron, and its effect can be clearly illustrated 
 by placing a compass inside a conning-tower with 12" plating, 
 the directive power would be 0'2 (assuming the earth's 
 horizontal force to be 1), if the plating were increased in 
 thickness until it was one-tenth the diameter of the 
 eonning-tower, the directive power of the needle would be 
 destroyed. 
 
 In a general magnetic sense, the cause may be illustrated 
 in this manner — if the conning-tower were made of wood 
 (Fig. 10), the lines of force would be parallel and equidistant, 
 
 Fig. 10.* Fig. 11.* 
 
 the magnetic field being uniform, and the compass needle 
 unaffected; but (Fig. 11) substitute soft iron for wood, th^ 
 magnetic field is no longer uniform, lines of force are no 
 longer parallel and equidistant, through the lines of force 
 travelling through the iron instead of the air, the former 
 being more permeable, i.e., offering less resistance to the lines 
 of force, f 
 
 In an iron ship, the decks, bulkheads and beams (con- 
 tinuous iron) appear to attract the lines of force and lessen 
 the directive power of the needle, but detached masses 
 
 * In the south of England these lines would be inclined to the 
 horizontal plane at an angle of 67 ° 20 ' 
 
 t The placing of armour-plates around the ship's dynamos will 
 prohahly throw some additional light on this subject. 
 
 E 2 
 
52 THE MARINER'S COMPASS. 
 
 (quadrantal spheres)* increase the mean directive force, by 
 increasing it when north and south, and decreasing it by 
 half the augmented amount when east and west. 
 
 The earth's horizontal force in the south of England is 
 assumed to be 1,. The mean value of the horizontal force in 
 the ship for a complete circle in azimuth is called \ (lambda). 
 The proportion between the earth's and compass horizontal 
 force will be inversely as the square of the number of seconds 
 occupied by a needle in making the same number of vibra- 
 tions on shore, and when in the position of the compass. 
 In England a needle f usually makes ten vibrations in 
 nineteen or twenty seconds, supposing the needle to make 
 ten vibrations in twenty seconds on shore, and on being 
 placed in the compass pedestal makes the same number of 
 vibrations in twenty-eight seconds : 
 
 H' (Compass horizontal force) = H (Earth's horizontal 
 
 force) X §: 
 
 H'=l X ~ = 0-51, 
 
 Showing that in that direction of the ship's head the ship's 
 force is acting against the earth's force; but if the time of 
 making the same number of vibrations be decreased to 
 sixteen seconds, it would shew that the ship's and earth's 
 horizontal force were acting in conjunction, and H' would 
 be 1*5. H' (being at the compass), is composed of the 
 
 * See pages 41 and 43. 
 
 t The needle should be about three inches long, flat and pointed, with 
 a jewelled cap working on a pivot ; the latter should fit the pivot holder 
 in compass bowl. The needle is vibrated in the horizontal plane by means 
 of a weight placed to counteract the dip. It should be used in a closed 
 box or compass bowl to prevent the needle being disturbed by the motion 
 of the air, the arcs of vibration should be as small as convenient to 
 lessen the effects of friction and resistance of the air, and should not 
 exceed an arc of 45 ° at the commencement. The horizontal components 
 may be compared by the deflections produced by a magnet placed at right 
 angles to the needle. 
 
C {PORTION OF B). 53 
 
 horizontal components of the earth, and the ship's permanent 
 and induced magnetism in the direction in which the needle 
 points {i.e., the compass north), to ascertain H' for magnetic 
 north, the observed horizontal force must be multiplied 
 by the cosine of the deviation. X can be ascertained by- 
 observations on four or more equidistant true or magnetic 
 
 H' 
 azimuths, from the mean of several values of fr cos S, or the 
 
 mean of the forces at N and S. 
 
 In the previous chapter, the causes of semi-circular error 
 have been stated ; and in the present one that the semi- 
 circular error can be represented by the co-efficients B and C, 
 and that B is composed of (P), the force acting in the fore 
 and aft line due to the permanent magnetism of the ship, its 
 effect changing inversely as the earth's horizontal force and 
 (c) due to vertical induction in soft iron, varying with the 
 
 P 1 
 tangent of the dip* B = (c x tan. dip + xf ) x r 
 
 At the magnetic equator (no dip) c wOl be nil, and if 
 B and C were corrected by permanent magnets, the change 
 thereafter, when proceeding to higher magnetic latitudes 
 would have to be corrected by Flinders bar. For a similar 
 reason if it were found that a Flinders bar was not required 
 in England, it would be required still less in the principal 
 portions of the globe visited by men-of-war, where the 
 horizontal force is greater and the dip less than in England, 
 the exceptions being Norway and Iceland on the Home, the 
 neighbourhood of Behring Strait on the Pacific station, and 
 that portion of the North American station north of New 
 York, with especial reference to the Gulf of St. Lawrence 
 and the coast of Labrador. A mathematical expression 
 giving the component parts of B, viz., P and c, without 
 
 * c is marked + when the nortli end of the needle is drawn towards 
 the bow, and — when towards the stern. The FUnders bar in the former 
 case must be placed abaft, and in the latter before the compass to counter- 
 act the error. The change of polarity in vertical error which takes place 
 in crossing the magnetic equator is gradual not abrupt. 
 
54 TEE MABINES'S COMPASS. 
 
 change of geographical position is known,* but fot the 
 reasons given later on,t it ia considered an uildefeiraljld 
 method. The component parts of B Can be oBtdiif^d, 
 provided B has been ascertained in two different magnetib 
 latitudes, where the dip| and HJ horizontal force are 
 known : X is generally known before the ship's departure 
 from England, or must be ascertained according to the 
 directions given above. § The more widely separated the 
 magnetic latitude (i.e., the dip), the better the results (60° 
 would give good results). The magnets correcting, B must 
 remain unchanged in number and position between the 
 observations, or the uncorrected value of the deviation must 
 be known on the east and west poiilts at the two observing 
 stations ; and in the mercantile marine, the cargo if magnetic 
 must remain undisturbed. 
 
 P c 
 
 ^ + H tan. dip j^ = BH 
 
 ^ + H' tan. dip' I = B'H' 
 
 The proper signs prefixed to the dip || 
 
 Taking in the south of England B = — -074 
 Do. do. Cape B = - -136 
 
 c at the first named place would require a correction qf 2° * 7, 
 and at the latter — 1°'6, the amount to be corrected by 
 
 c 
 
 Flinders bar for any intermediate position would be r tan, 
 
 dip ; the equivalent length of Flinder's bar for values of c, ^ 
 are based on a bar 3 inches thick and 9" from the centre of 
 the compass. 
 
 Supposing no dominant magnetic foroe, owing to its 
 proximity, to control the needle, the influence of the general 
 
 * Admiralty Manual, page 157. 
 
 t See page 55. Heeling error co-efficient. 
 
 I Given with sufficient accuracy for this purpose in the magnetic chairts 
 at the end of this work. 
 
 § Page 52. 
 
 II See Dip, page 16. t See Table III. 
 
EEELINQ ERBOB. 55 
 
 magnetism of the ship to the induced magnetism in vertical 
 soft iron, will be P 0*75, or 0'9 toe 0*25 or 0-1. 
 
 Heeling Error Co-effioient. — It is desirable before 
 proceeding to describe the heeling co-efficient to reiterate 
 what is practically done or attempted when the heeling 
 magnets are placed in position, and the extent of the 
 correction. To correct a compass, already corrected for 
 quadrantal and semi-circular errors, for heeling error is to 
 make* "the vertical magnetic force at the place of the 
 compass on board to bear the same ratio to the vertical 
 magnetic force on shore, that the horizontal force on board 
 bears to the horizontal force on shore," the extent and value 
 of the heeling error correction will be best judged by the 
 following opinions. The late Sir G. B. Airy writes : " There 
 appears to be no safe way of determining the amount of the 
 eflFect of heeling, except by making the ship to heel and 
 observing how much the compass is affected." Lord Kelvin 
 remarks, " Although the heeling magnet is very useful in 
 bring the heeling error within moderate bounds, the mariner 
 must constantly be on his guard to determine the heeling 
 error by observation,' and allow for it when the ship heels 
 over, even though he may know his compass to be correct 
 when the ship is upright." Captain Mayes (formerly Super- 
 intendent of Compasses) in the Journal of the E. U. S. I. 
 states : " The correction of the heeling error is unsatisfactory, 
 and the formula given by Mr. Archibald Smith does not 
 appear to be correct," and Captain Mayes' statement was 
 corroborated by a tabulated form f of the compass heeling 
 errors of H.M.S. " Orlando" as found by actual experiment 
 and those by calculation. As previously stated, the heeling 
 error I is mainly composed of three component parts, 
 (l) Vertical induction in transverse soft iron, represented 
 
 * Instructions for the adjustment of Sir William Thomsonjs (Lord 
 Kelvin) patent compass. 
 
 t Given by Staif Commander (now Captain) Creak, Superintendent of 
 Compasses. 
 
 X See Heeling error, page 43. 
 
56 TEE MABINEES COMPASS. 
 
 by D and \. (2) Vertical induction in vertical soft iron, and 
 (3) Vertical permanent magnetism of ship, they are repre- 
 sented by the co-eflSeient [i (mu). The majority of ships in 
 England are built with their heads to the southward,* and 
 have .their standard compasses forward, so that (3) will act 
 downwards, and combine with the soft iron of the ship. 
 
 The co-efficient ju. is the mean value of the vertical 
 force at the needle to that of the earth, and is calculated 
 from vibrations or by statical measurement. Let Z' be the 
 vertical force of earth and ship on any azimuth 9, and Z the 
 earth's vertical force : 
 
 Z' 
 
 fji (for any oi 6) = ^ 
 
 Z' 
 
 fi can be attained with the ship's head east or west ; with -p- 
 
 Z' 
 
 on any two azimuths 180° apart ; or the mean of y on any 
 
 Z' 
 
 number of equidistant azimuths. y can be ascertained 
 
 (except where the vertical force is small), by landing in a spot 
 free from any local attraction, a freely suspended dipping 
 needle, that is, the instrument must be level ; with the 
 needle's face in the magnetic meridian, the oscillations being 
 at right angles to the meridian, the needle when stationary 
 will be vertical^ the horizontal component of the earth's force 
 acting on the needle will be nil. Cause the needle to 
 vibrate ten times through an arc of 15° to 20° from each 
 side of the vertical, note the time, or the mean of times of 
 several sets, calling it ^.f Suspend the same needle in the 
 binnacle, note the time of vibrations or the mean of the times, 
 
 * Except ships built in private yards, whicli are located on the north 
 side of rivers. 
 
 t By placing a small semi-circular rod of wood, with its flat side 
 downwards, across the top of the binnacle or pedestal, slightly notched in 
 the centre for the chain suspending the instrument, the latter will level 
 itself, and the observations less likely to be disturbed. The resting points 
 of the wood on the binnacle top should be scored for a similar reason. 
 
EEEIING EBB OB INSTBUMENT. 57 
 
 calling it t', as the forces will vary inversely as the square of 
 
 f iZ 
 
 the times of vibration 7-2 = p, ; Z being considered as 1. 
 
 By statical 'measurement : Heeling error instrument 
 consists of a brass case fitted with a spirit level, one side 
 being of glass and on it is marked a horizontal line ; the 
 instrument stands on a flat foot for shore observations, where 
 when used, it should be placed, levelled, on a pedestal nqt 
 less than three feet from the ground. For ship observations, 
 it is suspended by a chain and hook in the top of the instru- 
 ment, from a wooden rod laid across the binnacle top in the 
 manner previously described, and the instrument should 
 level itself, but any small alteration of level can be made by 
 a careful movement of the links of the chain. The needle is 
 cylindrical, with pointed ends, the red or north seeking end is 
 marked by a circle, it is graduated, every fifth division being 
 marked by a dot, the needle is fitted with a sliding 
 aluminium weight or weights ; if two be used, they must 
 touch one another and be regarded for purposes of measure- 
 ment as one ; it is mounted on a transverse axis, knife-edged 
 on its under side, and the needle when not required for 
 observation is capable of being lifted on brass bearers by a 
 milled screw at the back and bottom of the brass case. 
 
 On shore, at a spot free from local attraction, the instru- 
 ment being levelled, is turned in azimuth until the marked end 
 of the needle is directed to the north. Place the aluminium 
 weight or weights on the uppermost end (unmarked in north 
 magnetic latitudes) until the needle is horizontal, i.e., coin- 
 ciding with the horizontal line marked on the glass face, the 
 distance (a) is noted from axis of suspension. 
 
 Remove the compass to the ship, her head * being east 
 or west magnetic, suspend the instrument to the wooden 
 rod as previously described, with its needle in the same 
 
 * For the reasons given on last page of Chapter IV. The ship is 
 supposed to be on even beam — if heeled the ship's head should be north or 
 south magnetic. 
 
58 TEE MASINEB'S COMPASS. 
 
 horizontal plane to that lately occupied by the needles of 
 
 the compass card. Turn the instrument in azimuth until the 
 
 red end of the needle points to magnetic north (nearly), 
 
 again make the needle horizontal by the sliding weight 
 
 or "weights, and the distance (6) noted from the axis of 
 
 suspension 
 
 Z:Z'::a:b. 
 
 The method of vibrations is generally adopted to 
 ascertain the values of (2 and 3) the component parts of ju. or , 
 the amount of heeling error ; but for mechanical correction 
 of the heeling error the method by statical measurement is 
 preferable. 
 
 fi, as already mentioned, represents the combined effects 
 of (2 and 3) portions of the heeling error, but (1) the 
 portion arising from vertical induction in transverse soft 
 iron is composed of \ and D. 
 
 To ascertain the heeling error by vilprations : 
 
 t'= * 
 
 ^/ X (1 + D).* 
 
 To ascertain the heeling error by heeling error instru- 
 ment : 
 
 5 = Xa(l + D).* 
 
 The heeling error can be corrected at sea by means of 
 the heeling instrument, provided the value of (a) is known 
 at a zero station of high vertical force, such as the British 
 Naval Sefiports. 
 Let a = Number of scale divisions (30) at zero station 
 6 = • do. at sea station (re- 
 
 quired) 
 Z = Vertical force at zero station (2*36) | -p, , 
 Z'= do. at sea station (0*50) j 
 
 tX(l +D) = 0-908 
 
 * The quadrantal error is corrected. 
 
 t X is practically constant for navigational purposes, increasing with 
 time but decreasing with heat. D also is practically constant for naviga- 
 tional purposes, but decreases with time. 
 
HEELING iBB OR INSTR UMENT. 59 
 
 h Z' 
 
 -=^xX(l+D) 
 
 . axZ'xX(l+D) 30 X 0-50 X -908 
 
 * = Z 2^6 = ^-^6 
 
 a = 5.76 tlie number of scale divisions to which the 
 heeling error instrument must be set. 
 
 The Heeling Co-efficient, if known, on the north or 
 south points (points of maximum disturbance) can be found 
 for any other azimuth by multiplying it by the cosine of the 
 azimuth, measured from the points of niaximum disturbance. 
 
60 THE MARINER'S COMPASS. 
 
 CHAPTER VI. 
 
 We now proceed to the mechanical correction of the compass ; 
 first the order of the correction, and the reasons for their 
 sequence. It is hardly necessary in the present day to state 
 the reasons for correcting a compass, but the principal ones 
 are equalising the directive force of the needle* for all points 
 of the compass and increasing the steadiness of the compass 
 in a seaway, thus preventing sluggishness on certain ones ; 
 also to cause the angular movement in azimuth, and ship's 
 head to nearly coincide,! a great assistance in fleet tactics, 
 and taking rough bearings at night. A compass should be 
 corrected in the following order — (1) The quadrantal error, 
 (2) the heeling error, and lastly the semi-circular error. A 
 ship is never of such a diverse type from any previous one, 
 but that a rough approximation can be made of her probable 
 quadrantal error (D). The spheres when in place increase 
 the mean directive force of the needle J and the heeling error 
 caused by vertical induction in soft iron is largely corrected 
 by the spheres. Heeling error is next corrected, if the ship 
 were not on even beam when correcting (C), a heeling error 
 is introduced which has its maximum on north and south 
 points. The vertical magnets and ends of the fore and aft 
 ones should not be capable of inducing magnetism in 
 Flinders bar (for correction of B (c) ). 
 
 * The frictional error is, roughly, inversely proportionate to the 
 directive force. 
 
 t See table of the large deviations of the " Trident " and " Achilles " in 
 Admiralty Manual. 
 
 i A case occurred in which a compass kept at east for an entire 
 revolution in azimuth, until the spheres were placed in position. 
 
MECHANICAL CORRECTION. 61 
 
 If a Flinders bar be used, it should be in place before the 
 vertical magnets are finally adjusted. 
 
 To Adjust. — Place quadrantal spheres in place, according 
 to previous experience ; and as ship proceeds down a river or 
 to a buoy, diminish the semi-circular error if known to exist. 
 On being made fast to a buoy,* swing on the cardinal and 
 inter-cardinal points — ascertain D, to correct a -t- Df move 
 spheres closer to compass or use larger spheres, for — D 
 increase the distance or use smaller spheres. If the semi- 
 circular error is found to be large, it is better to adjust a 
 second time before correcting the heeling error, supposing if 
 it were necessary, that it has been done ; the Heeling error 
 is corrected, if by the method of vibrations, in moving the 
 can containing the vertical magnets upwards or downwards 
 by its chain, or by increasing or decreasing the power and 
 
 number of the magnets until t' = ^ , as explained in 
 
 the previous chapter. By Heeling error instrument with 
 the ship's head east or west,| move the can in a similar 
 manner until 6 = X a (1 -f D). § 
 
 Should a vessel proceed to sea in north magnetic 
 latitudes, unadjusted for heeling error and the compass in 
 consequence becomes unsteady, the heeling error may be 
 approximately corrected, by placing the ship as nearly as 
 practicable on a north or south course, noting as the ship 
 heels over, whether the north (red) end of the compass is 
 attracted to the windward side of the ship. 
 
 * As previously explained, a ship can be steamed round in a circle, 
 according to local circumstances. 
 
 •j" See Table IIa. It is understood that the needles are of such small 
 directive power, that practically they do not produce induced force in the 
 spheres ; with large needles the magnetic power of the spheres would be 
 increased, and the power would vary with the magnetic latitude. Single 
 needle compasses should not be used with spheres, as they cause an octantal 
 error. 
 
 X For the reasons given in last page of Chapter IV. and footnote page 57. 
 
 § With heeling error instruments supplied to H.M. ships, the value of 
 (X (1 -I- D)) is usually given in MS. for the principal compasses. 
 
62 THE MARINERS COMPASS. 
 
 If SO, the error should be corrected by magnet or magnets 
 red marked ends up ; if attracted to leeward side, by magnet 
 or magnets with their blue marked ends up— raising or 
 lowering the can containing the magnets, until a position is 
 reached, where the compass is (roughly) undisturbed by the 
 heeling of the ship. 
 
 On crossing the magnetic equator and proceeding to 
 higher magnetic latitudes, it may be necessary to reverse 
 the position of the poles of the magnets in the can. 
 
 The can's chain should, be carefully secured to prevent it 
 from slipping, to enable a navigating oflScer to know this has 
 not occurred, a link of the chain is simply marked by twine ; 
 between the bottom of the can and the bottom of the well 
 bored for its reception, should be three inches, after the 
 distance by previous experiment has been found ; as in many 
 compasses, for instance Lord Kelvin's (which is usually 
 adopted as a standard in H.M. Navy), the method of sus- 
 pension allows a lateral movement of the compass ; so when 
 the ship heels, the centre of the can and the centre of the 
 card are no longer in the same vertical plane and the heeling 
 error is over corrected ; it is therefore usual to prevent over- 
 correction by lowering the can one and a half inches* below 
 the amount obtained by instrumental means. 
 
 With the number and size of magnets usually employed 
 in heeling error correction, no serious disturbance from this 
 cause need be apprehended until the vertical magnets 
 approach within two feet of the compass cards, but should 
 any doubt exist it is better to practically test it, by inclining 
 the pedestal on a calm, jBne day. 
 
 To correct the Semi-circular deviation with the ship's 
 head magnetic North ; observe the compass-bearing of an 
 object, the magnetic-bearing of which is known. 
 
 (a) If the deviation be easterly ( + ) and the red (north) end 
 
 * Should 14-moli magnets be employed, the IJ- inches must be in« 
 creased to 2J. . 
 
MECHANICAL CORRMCTION. 63 
 
 of the needle is drawn to starboard, increase in the 
 transverse holes of the pedestal the number of 
 magnets with red ends to starboard, or place them 
 nearer the card, until the magnetic and compass 
 bearings coincide, and there is no deviation. 
 
 (6) If the division be westerly ( — ) and the red (north) 
 end of the needle is drawn to port, the magnets 
 must be placed with their red ends to port or 
 decrease the number of magnets with red ends 
 to starboard or remove them farther away from 
 the card. 
 
 (c) Or with ship's head magnetic South ; if- the deviation be 
 westerly ( — ) and the red (north) end of the needle 
 is drawn to starboard, in the transverse holes of the 
 pedestal, increase the number of magnets with red 
 to starboard or place them nearer the card. ' ' 
 
 {d) If the deviation be easterly ( + ) and the red (north) end 
 is drawn, to port, the magnets muSt'be placed with 
 their -red ends to port, or decrease the number of 
 magnets with red ends to starboard, or remove them 
 farther away from the card. 
 
 If there be no deviation, the co-efficient (C) repre- 
 senting the transverse magnetic force' in the ship has 
 been corrected. 
 
 {e) With the ship's head magnetic East ; if the observed 
 deviation be easterly ( + ) and th« red (north) end of 
 the needle is drawn towards the bow, in the fore and 
 aft holes of the pedestal, increase the number of 
 magnets with red ends forward or decrease their 
 distance from the card — the number of magnets in 
 port and starboard fore and aft holes should be equal 
 in number and strength. 
 
 (/) If the observed deviation be westerly ( — ) and the red 
 (north) end of the needle is drawn towards the stem, 
 the magnets must be placed with their red ends aft, 
 or the magnets with their red ends forward must be 
 
64 THE MARINER'S COMPASS. 
 
 reduced in number or their distances increased from 
 the card. Or with Ship's head Magnetic West. 
 
 (g) If the observed deviation be westerly ( — ) and the red 
 (north) end of the needle is drawn towards the bow 
 increase the number of magnets with red ends forward 
 or decrease their distance from the card. ' 
 
 (A) If the observed deviation be easterly ( + ) and the red 
 (north) end of the needle is drawn towards the stern, 
 the magnets must be placed with red ends aft, or the 
 magnets with their red ends forward must be reduced 
 in numbers, or their distances increased from the card. 
 
 In the above corrections, the principles of like poles repel 
 one another is brought into practice. 
 
 If there be no deviation, the co-efficient B, representing 
 the fore and aft magnetic force in the ship, has been corrected. 
 Should it bei considered desirable to correct that portion of 
 B (c) by the Flinders bar, and the knowledge (special to the 
 ship), with regard to the relative proportions of P and c are 
 unknown ; correct ^^^ to ^ of B by Flinders bar, the latter 
 being raised two inches above the level of the compass card. 
 
 A, the constant deviation (real), should be small in 
 amount in a properly-placed compass, and remain uncor- 
 rected, as it is constant for all courses, 
 
 E should be small in amount in a properly-placed com- 
 pass, and can be corrected, if desirable, as previously stated,* 
 by placing the spheres at such an angle {/3) with the athwart- 
 
 E 
 ship line that tan 2 /8 will equal j^. 
 
 The ship should then be swung, and the deviations 
 tabulated. The maximum deviation in a well-placed com- 
 pass should be less than 2°. 
 
 General Eemarks and Cautions. — On a navigating 
 officer joining a ship, he should examine the compass pedes- 
 tal, that it is properly secured to the deck : (1), the centre 
 
 * See page 48. 
 
GENERAL REMARKS AND CAUTIONS. 65 
 
 of the quadrantal spheres and compass needles should be in 
 the same horizontal plane, whether the latter are athwart- 
 ships, or at angle to the athwartship line ; (2),* a vertical 
 plane, passing through the centre of the compass needles, 
 should also pass through the centre of the horizontal plane of 
 the magnets ; (3), the spheres and magnets are correct with 
 regard to position, size and number, according to the deviation 
 table (if supplied) ; (4), the compass bowl moves freely in the 
 gimbals ; that the screws securing the glass verge are not 
 worn, as they are made with a special thread, rendering their 
 replacement difl&cult after leaving a dockyard, as no screw- 
 . driver is supplied, the one in the station pointer's box will be 
 found useful ; (6), that the lubber's line is in, or parallel to, 
 the fore and aft line ; to ascertain this for the Standard, the 
 ship should be on even beam ; set the index of the base plate 
 carrying the azimuth mirror at 180°, reflect the centre of any 
 object in the fore and aft line on to the lubber's point, if 
 they agree, the lubber's line is correct, Otherwise, turn the 
 pedestal in azimuth until they do agree ; if the object be 
 abaft, reverse the bowl in the gimbals, so that the lubber's 
 point will be aft ; (6), that one of the faces of the Azimuth 
 mirror when iu use is parallel to the face of the glass verge 
 by observing if a bearing remains constant when the object is 
 reflected to thfe graduation of the card, and when the gradua- 
 tion is reflected to the object ; if it difi'ers, the error can be 
 ascertained and allowed for ; (7), the key of the door of the 
 pedestal in which the correcting magnets are placed should 
 be in charge of the navigating officer, and any alteration in 
 number, of size or position, should be entered by him in 
 the compass journal ; (8), the minimum distance between 
 binnacles should be 41 feet. 
 
 Compass Cards should be stowed with unlike poles to- 
 gether, the numbers on the outside of the compass card- 
 boxes are so marked, that this is carried out by placing the 
 numbers above one another. The cards should be placed in 
 the boxes with the engraved face downwards,* the north 
 * Refers to Thomson's cards and fittings only. 
 
 P 
 
66 THE MARINER'S COMPASS. 
 
 end of the needle at the (N) end of the box, with the centre 
 of the card over the wooden stud, and the thread attached 
 to north or zero point should be placed between the two 
 metal studs. The pivots and caps should be examined and 
 shipped at convenient opportunity to see the card works 
 freely.* 
 
 To ship a Thomson's card, the pivot and cap are taken 
 from the triangular-shaped partition in the compass card- 
 box ; the pivot should be forced well home in the holder at 
 the bottom of the bowl ; place the centre of the card over 
 the cap — the latter with its rounded end up. Lift the card 
 by the upper portion of the cap, placing it carefully on the 
 pivot, with the card approximately in its correct magnetic 
 position with regard to the ship's head. 
 
 The Magnets supplied to the service are cylindrical, 
 with painted halves of red and blue (as suggested by the 
 late Sir G. B. Airy), the red to mark the end possessing the 
 same description of magnetism as the earth's south polar 
 regions, and blue to mark the end possessing the same 
 description of magnetism as the earth's north polar regions. 
 The majority are f 9" x f " and 9" x ^^" in size, and, except 
 for the purpose of correcting heeling error, are always placed 
 broadside on the compass,^ if the distance and power of a 
 magnet be constant, the force of a magnet end on is double 
 that when broadside on. The spare magnets of the same 
 size should be kept in bundles, with opposite poles together, 
 and placed in the compass locker and not in a drawer of a 
 chart-house, in proximity to a compass.§ The Controller's 
 department state -that four feet should be the minimum 
 
 * See remarks on manoeuvring compass on next page. 
 
 t For correcting heeling error fourteen inch magnets are occasionally 
 employed. 
 
 t An exception was the case of the " Undaunted " (old), a wooden frigate 
 with iron masts ; and exceptions may occur when dealing with vertical iron 
 immediately below the compass. 
 
 § A merchant ship has been known to have had cushions with steel- 
 wire springs on the lockers in a wheel-house, to the confusion of the 
 adjuster, until the source of the error was discovered. 
 
OENERAL BEMABKS AND CAUTIONS. 67 
 
 distance between compasses and chronometers ; this should 
 be borne in mind if the latter are removed when preparing 
 for action, and that the intensity of magnetisation of the 
 needles in a liquid compass and a Thomson are widely 
 different. 
 
 In practice, when swinging, it is usual to observe alter- 
 nate points, but including the cardinal and inter-cardinal 
 ones ; and the ship should be kept steady for a few minutes 
 on each point observed, as a ship takes some little time to 
 acquire the magnetism due to that point, and vertical iron 
 is often more sluggish in parting with its induced magnetism 
 than horizontal. The following precautions should be taken 
 in swinging, observing that the ship should always be swung 
 by the standard compass : (1), the magnetic bearing of an 
 additional distant object should be known, in case the 
 original one becomes obscured by masts, funnels, etc. ; (2), 
 in reflecting with the azimuth mirror, avoid unequal pres- 
 sure on opposite sides of the compass bowl, more especially 
 in a seaway, or blowing hard ; (3), if reflecting a celestial 
 body, the smaller the altitude the better ; the ordinary limit 
 should be 30°, to be increased to 40° under extraordinary 
 circumstances ; (4), if a shadow pen be used, it should be 
 straight and upright.* Finally, the reader is recommended 
 to again peruse the opening paragraphs of Chapter I. 
 
 * A manoeuvring compass should be kept closely adjusted by the 
 navigating officer, and affords him a ready means of obtaining a practical 
 knowledge of the adjusting. Should a compass become sluggish, due, as 
 far as can be judged, from other than magnetic disturbance, the cap and 
 pivot should be changed ; a trifling imperfection in either of them may 
 be coincident with the moment of the compass. The average period of a 
 Thomson's card varies from thirty seconds for a ten-inch card, to thirteen 
 and a half seconds for a four-inch one. 
 
68 
 
 THE MARINER'S COMPASS. 
 
 TABLE \. 
 
 For the computation of the Coefficients, B, G, D, E^ 
 
 Products of Arcs Multiplied by the Sines of the Khumbs. 
 
 ARCS. 
 
 Sin. 1H°. 
 
 Sin. 22i°. 
 
 Sin. 33i° 
 
 Sin. 46°. 
 
 Sin. 66i°. 
 
 Sin. 67i°. 
 
 Sin. YSJ". 
 
 
 
 
 
 
 
 
 
 
 AKCS. 
 
 'ijat Sin. 
 
 •19609. 
 
 ■38268. 
 
 •5566?. 
 
 '70?10. 
 
 •83U7. 
 
 •92388. 
 
 •98078. 
 
 
 O / 
 
 1 
 
 / 
 
 o / 
 
 O 1 
 
 O 1 
 
 / 
 
 o / 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 10 
 
 2 
 
 4 
 
 6 
 
 7 
 
 8 
 
 9 
 
 10 
 
 10 
 
 20 
 
 4 
 
 8 
 
 11 
 
 14 
 
 ,0 17 
 
 18 
 
 20 
 
 20 
 
 30 
 
 6 
 
 11 
 
 17 
 
 21 
 
 25 
 
 28 
 
 29 
 
 30 
 
 40 
 
 8 
 
 15 
 
 22 
 
 28 
 
 33 
 
 37 
 
 39 
 
 40 
 
 50 
 
 10 
 
 19 
 
 28 
 
 35 
 
 42 
 
 46 
 
 50 
 
 50 
 
 t 
 1 
 
 12 
 
 23 
 
 33 
 
 42 
 
 50 
 
 55 
 
 59 
 
 1 
 
 1 10 
 
 14 
 
 27 
 
 39 
 
 49 
 
 58 
 
 1 5 
 
 1 9 
 
 1 10 
 
 1 20 
 
 16 
 
 31 
 
 44 
 
 57 
 
 1 7 
 
 1 14 
 
 1*18 
 
 1 20 
 
 1 30 
 
 18 
 
 34 
 
 50 
 
 1 4 
 
 1 15 
 
 1 23 
 
 1 28 
 
 1 30 
 
 1 40 
 
 20 
 
 38 
 
 56 
 
 1 11 
 
 1 23 
 
 1 32 
 
 1 38 
 
 1 40 
 
 1 50 
 
 22 
 
 42 
 
 1 1 
 
 1 18 
 
 1 32 
 
 1 42 
 
 1 48 
 
 1 50 
 
 ■n 
 
 2 
 
 23 
 
 46 
 
 1 7 
 
 1 25 
 
 1 40 
 
 1 51 
 
 1 58 
 
 2 
 
 2 10 
 
 25 
 
 50 
 
 1 12 
 
 1 32 
 
 1 48 
 
 2 
 
 2 8 
 
 2 10 
 
 2 20 
 
 27 
 
 54 
 
 1 18 
 
 1 39 
 
 1 56 
 
 2 9 
 
 2 17 
 
 2 20 
 
 2 30 
 
 29 
 
 57 
 
 1 23 
 
 1 46 
 
 2 5 
 
 2 19 
 
 2 27 
 
 2 30 
 
 2 40 
 
 31 
 
 1 1 
 
 1 29 
 
 1 53 
 
 2 13 
 
 2 28 
 
 2 37 
 
 2 40 
 
 2 50 
 
 33 
 
 1 5 
 
 1 34 
 
 2 
 
 2 21 
 
 2 37 
 
 2 47 
 
 2 50 
 
 3 
 
 35 
 
 1' 9 
 
 1 40 
 
 2 7 
 
 2 30 
 
 2 46 
 
 2 57 
 
 3 
 
 310 
 
 37 
 
 1 13 
 
 1 46 
 
 2 14 
 
 2 38 
 
 2 56 
 
 3 6 
 
 3 10 
 
 3 20 
 
 39 
 
 1 17 
 
 1 51 
 
 2 21 
 
 2 46 
 
 3 5 
 
 3 16 
 
 3 20 
 
 3 30 
 
 41 
 
 1 20 
 
 1 57 
 
 2 29 
 
 2 55 
 
 3 14 
 
 3 26 
 
 3 30 
 
 3 40 
 
 •0 43 
 
 1 24 
 
 2 2 
 
 2 36 
 
 3 3 
 
 3 23 
 
 3 36 
 
 3 40 
 
 3 50 
 
 45 
 
 1 28 
 
 2 8 
 
 2 43 
 
 3 11 
 
 3 33 
 
 3 46 
 
 3 50 
 
 4 
 
 47 
 
 1 82 
 
 2 13 
 
 2 50 
 
 3 20 
 
 3 42 
 
 3 55 
 
 4 
 
 4 10 
 
 49 
 
 1 36 
 
 2 19 
 
 2 57 
 
 3 28 
 
 3 51 
 
 4 5 
 
 4 10 
 
 4 20 
 
 51 
 
 1 40 
 
 2 24 
 
 3 4 
 
 3 36 
 
 4 
 
 4 15 
 
 4 20 
 
 4 30 
 
 53 
 
 1 43 
 
 2 30 
 
 3 11 
 
 3 45 
 
 4 10 
 
 4 25 
 
 4 30 
 
 4 40 
 
 55 
 
 1 47 
 
 2 36 
 
 3 18 
 
 3 53 
 
 4 19 
 
 4 35 
 
 4 40 
 
 4 50 
 
 57 
 
 1 51 
 
 2 41 
 
 3 25 
 
 4 1 
 
 4 28 
 
 4 44 
 
 4 50 
 
 5 
 
 59 
 
 1 55 
 
 2 47 
 
 3 32 
 
 4 9 
 
 4 37 
 
 4 54 
 
 5 
 
 5 10 
 
 1 1 
 
 1 59 
 
 2 52 
 
 3 39 
 
 4 18 
 
 4 46 
 
 5 4 
 
 5 10 
 
 5 20 
 
 1 2 
 
 2 3 
 
 2 68 
 
 3 46 
 
 4 26 
 
 4 56 
 
 5 14 
 
 5 20 
 
 5 30 
 
 1 4 
 
 2 6 
 
 3 3 
 
 3 53 
 
 4 34 
 
 5 5 
 
 5 24 
 
 5 30 
 
 5 40 
 
 1 6 
 
 2 10 
 
 3 9 
 
 4 
 
 4 43 
 
 5 14 
 
 5 34 
 
 5 40 
 
 5 50 
 
 1 8 
 
 2 14 
 
 3 14 
 
 4 8 
 
 4 51 
 
 5 23 
 
 5 43 
 
 5 50 
 
 6 
 
 1 10 
 
 2 18 
 
 3 20 
 
 4 15 
 
 4 59 
 
 5 33 
 
 5 53 
 
 6 
 
TEE MARINERS COMPASS. 
 
 69 
 
 Tables II., Ila., and III., with the Magnetic Charts, are taken from 
 Admiralty Manual {edition 1893). 
 
 TABLE II. 
 
 QPADRANTAL DEVIATION. COEEEOTION BY SPHEBES. 
 
 Distances of Sphe/res to correct given values of Goeffkient D. 
 10" Thomson Binnacles, brass brackets. 
 
 Position of 
 
 Amount of D that will be corrected with different sized Spheres. 
 
 Spheres. 
 
 6" Spheres. 
 
 6" Spheres. 
 
 ?" Spheres. ' 
 
 8i" Spheres. 
 
 10" Spheres. 
 
 12" Spheres. 
 
 
 / 
 
 ( 
 
 I 
 
 / 
 
 1 
 
 / 
 
 Close up* 
 inches. 
 
 2 20 
 
 3 
 
 4 39 
 
 6 20 
 
 8 
 
 10-36 
 
 Set out • 1 
 
 2 17 
 
 2 55 
 
 4 32 
 
 6 10 
 
 7 46 
 
 
 •2 
 
 2 13 
 
 2 50 
 
 4 25 
 
 6 
 
 7 33 
 
 
 •3 
 
 2 10 
 
 2 45 
 
 4 20 
 
 5 50 
 
 7 20 
 
 
 .4, 
 
 2 6 
 
 2 41 
 
 4 15 
 
 5 42 
 
 7 10 
 
 
 •5 
 
 2 2 
 
 ^ 2 38 
 
 4 10 
 
 5 35 
 
 7 
 
 
 ■6 
 
 2 
 
 2 34 
 
 4 3 
 
 5 28 
 
 6 49 
 at 
 
 
 •7 
 
 1 57 
 
 2 30 
 
 3 57 
 
 5 20 
 
 extreme. 
 
 
 •8 
 
 1 53 
 
 2 26 
 
 3 51 
 
 5 12 
 
 
 
 ■9 
 
 1 50 
 
 2 23 
 
 3 45 
 
 5 7 
 
 
 
 1-0 
 
 1 48 
 
 2 19 
 
 3 40 
 
 5 
 
 
 
 1-1 
 
 1 45 
 
 2 16 
 
 3 34 
 
 4 53 
 
 
 
 1-2 
 
 1 42 
 
 2 12 
 
 3 29 
 
 4 47 
 at 
 
 
 
 1-5. 
 
 1 36 
 
 2 3 
 
 3 13 
 
 extreme. 
 
 
 
 2-0 
 
 1 26 
 
 1 49 
 
 3 3 
 
 at 
 
 
 
 
 2-5 
 
 1 15 
 
 1 36 
 
 at 
 
 extreme. 
 
 
 
 
 3-0 
 
 1 4 
 
 extreme. 
 
 
 
 
 
 3-5 
 
 56 
 
 at 
 
 extreme. 
 
 
 
 
 
 ' 
 
 * Distance of centre of CompaBS to the oeaieet point of Spheres in the position 
 "close up" = 8'5 inches. 
 
70 
 
 THE MARINER'S COMPASS. 
 
 TABLE IlA.. 
 
 QUADEAITTAL DEVIATION. CoBEBCTION BY SPHERES. 
 
 Distances of Spheres to correct given values of Coefficient D. 
 6" Compass in Thomson Conning Tower Suspension, brass brackets. 
 
 Position of 
 
 
 Amount of D that will be corrected witii different sized Spheres. 
 
 Spbeiea. 
 
 5" Spheres. 
 
 6" Spheres. 
 
 7" Spheres. 
 
 8i" Spheres. 
 
 10" Spheres. 
 
 Close U 
 
 incl 
 Set out 
 
 p* 
 les. 
 
 1 
 
 O r 
 
 5 
 4 50 
 
 u / 
 
 7 35 
 
 7 18 
 
 10 50 
 10 26 
 
 o / 
 
 13 50 
 13 20 
 
 / 
 
 l(j 15 
 
 
 2 
 
 4 40 
 
 7 
 
 10 5 
 
 12 53 
 
 
 
 3 
 
 4 30 
 
 6 44 
 
 9-48 
 
 12 30 
 
 
 
 4 
 
 4 21 
 
 6 30 
 
 9 30 
 
 12 10 
 
 
 
 5 
 
 4 12 
 
 6 18 
 
 9 10 
 
 11 48 
 
 
 
 6 
 
 7 
 
 4 4 
 3 56 
 
 6 05 
 5 55 
 
 8 53 
 8 38 
 
 11 32 
 at extreme. 
 
 ' 
 
 
 8 
 
 3 48 
 
 5 45 
 
 8 22 
 
 
 
 
 9 
 
 3 40 
 
 5 35 
 
 8 8 
 
 
 
 1 
 
 
 
 3 32 
 
 5" 25 
 
 7 52 
 
 
 
 1 
 
 1 
 
 3 27 
 
 5 16 
 
 7 38 
 
 
 
 1 
 
 2 
 
 3 21 
 
 5 9 
 
 7 23 
 
 
 
 1 
 
 2 
 2 
 
 5 
 
 4 
 
 3 2 
 
 2 34 
 
 2 12 
 at extreme. 
 
 4 48 
 
 4 12 
 at extreme. 
 
 7 15 
 at extreme. 
 
 
 
 * Distance of centre of Compass to the nearest point of Spheres in the position 
 "olosenp" = 5'1 inches. 
 
 KoTE. — This Table will serve for 6" Binnacles if 30' be added to the Talnes here 
 leoorded. 
 
THE MARINER'S COMPASS. 
 
 TABLE III. 
 
 Semicieculae Deviation. 
 Correction of c, the changing pwrt of coefficient B, by Flinders bar. 
 
 Table of Equivalent lengths of Flinders Bar, to correct values of C. 
 
 Value of c. 
 
 •01 
 •02 
 •03 
 •04 
 •05 
 •06 
 •07 
 •08 
 •09 
 •10 
 •11 
 •12 
 •13 
 •14 
 •15 
 ■Ifi 
 
 Amonnt to be corrected 
 
 in sonth of England 
 where tan. dip = 2*33. 
 
 1 20 
 
 2 40 
 
 4 00 
 
 5 20 
 
 6 40 
 
 8 00 
 
 9 25 
 10 45 
 
 12 06 
 
 13 30 
 
 14 50 
 
 16 15 
 
 17 35 
 
 19 00 
 
 20 30 
 
 21 55 
 
 Length of bar 
 in Inches. 
 
 6^5 
 8-2 
 9-5 
 10^8 
 12^2 
 13^2 
 14^2 
 15^2 
 16^3 
 17^4 
 18-4 
 19-5 
 20-6 
 21^9 
 23 •! 
 24-4 
 
 Note. — These values have been computed for an average Flinders bar of malleable 
 iron, 3 inches in diameter, and made up of different lengths. If o be known, or can be 
 assumed, the equivalent length of bar can be at once placed in position. 
 
 When c has the sign +, on the after side of Binnacle. 
 
 When c has the sign — , on the fore side of Binnacle. 
 
 These values of c and equivalent lengths of bar are chiefly intended for a first 
 adjustment. 
 
 A, if unknown, may be assumed. 
 
 * For an iron or steel ship . . . ' 0'90. 
 
 * With armoured decks in above . . • . . . 0'85. 
 
 * Elementary Manual, page 1S6. 
 
LONDON: 
 FEINTED BY WILLIAM CLOWES AND spNS, Limited 
 
 STAMFORD STBEST AND CHARING CROSS. 
 
O o c5aoOoocf— "— — N N ni 
 
oLp-jo ^oojapa-y;'] - 
 
 en u> m o 
 
 ,n »n m tn 
 
 ooooo oood> d 
 
 m lO i/iiAiDior-c^ eo <n o — n •«f iS ^ ''^ 
 o b 6<ibboo66— — — ^— cj A