Introductory Observations.
It is sufficiently well known to the members of the Naval Institute that a change in the compass equipment of our ships of war has been gradually going on for a number of years past, until there has come to be a nearly complete substitution of the liquid compass for the air or dry compass previously in use.
The first introduction into our service of this kind of compass was about eleven or twelve years ago, in the form of a liquid boat-compass, the original models of which were imported compasses of English make, with flat porcelain cards.
The superiority of these instruments over the old and comparatively useless form of air boat-compass was soon evident and generally acknowledged; and yet, for various reasons (possibly from a little conservative prejudice), the new compass was only partially accepted, the outfit for boat equipment continuing for some time to consist partly of liquid and partly of air compasses.
Meanwhile, Mr. E.S. Ritchie, of Boston, who had long been distinguished for his intelligence and skill as a maker of the higher grades of philosophical instruments, finding it expedient, like many others at that time, to turn his attention to the fabrication of "war material" of some hind, solicited and obtained orders to make compasses for the Navy. It was thus that, while making some liquid boat-compasses after the English model, he was led to propose an important improvement, in the substitution of a buoyant card, with a pivot-pressure entirely at control, instead of the heavy flat disk otherwise in use.
This was the fruitful germ of Mr. Ritchie's idea, which, from the time he first put it into a practical form, has been in course of continual development, until the compass of the past year, although preserving its original distinctive features, is immeasurably the superior of its early predecessor in all the details of its construction.
Still, our naval authorities, while admitting the improvement, were far from precipitate in changing the compass equipment of our ships; and thus it happened that the new liquid compass was hardly brought into exclusive use, even for the steering-binnacle, until about five years ago, and for azimuth purposes only during the last two or three years. In reality, however, it is but few years since the first attempt was made to adapt the compass-card to azimuth observations by giving it a suitably-divided circle; and it is only within the past year that I have been satisfied the true construction of the compass-card had been finally reached.
It was thus by these cautious and tentative steps, and rather behind than in advance of established convictions, that the liquid compass came at length into general use in the Navy; and the time has arrived when we may, I think, with some propriety, put on record, in a public manner, some of the reasons which have appeared to justify a position in this regard so much at variance with the general practice of all other countries.
I propose, therefore, with your indulgence on this occasion, to present a few considerations relative to the principles which, as I am led to believe, should control the construction of the marine compass; and then, by way of application, to show wherein our practical conclusions have been, or are likely to be, justified by the particular type of compass now recognized in our naval service.
The subject is certainly not a hackneyed one; for, notwithstanding the considerable antiquity of the marine compass, in the general form by which it has been known to the nations of Western civilization—with volumes of history, popular description, and panegyric—I am not aware that a single attempt has hitherto been made to give a rational explanation of its magneto-mechanical action, or of the principles upon which its construction, as an instrument of observation, should depend. And yet the theory of this instrument rests upon a few simple considerations of certain established conclusions of science. The neglect in this particular may, perhaps, afford a sufficient explanation of what appeared, in my first acquaintance with the marine compass, as one of the most remarkable facts in its history—that, as commonly seen in the hands of navigators, it should be one of the rudest and most imperfect instruments for the accomplishment of one of the most important purposes among the practical pursuits of life.
There are three properties of the marine compass, so essential to its reliable action and convenient use that they may very properly be regarded as fundamental desiderata. They are:
MAGNET-POWER, SENSIBILITY, STEADINESS.
Let us consider, for a few moments, the precise significance of these terms in their present application ; the conditions that appear to be requisite for the most favorable realization of the properties which they represent; and how nearly they are likely to be satisfied by the present Navy compass.
I. MAGNET-POWER OF THE COMPASS.
The more or less complex system called a compass-card, alike with the simple magnetic needle, when balanced for motion in a horizontal plane, has a tendency to return to its position of equilibrium, or rest, whenever deflected from it, which may be called its moment of deflection.
The moment of deflection is equal to the moment of motive force less the moment of resisting force. The moment of motive force is composed of three factors: of which one is the magnet-power (or, perhaps, in more precise phrase, the magnetic moment) of the card; another is the directive force acting on the compass, meaning the whole exterior magnetic force acting in the horizontal plane of the card; and the third factor is the sine of the angle by which the zero-line of the card is deflected from its position of rest.
The moment of resistance is the moment of all the resistances, of whatever kind, to the motion of the card.
All these moments are referred to the point of the pivot as the centre of the compass and centre of moments.
In relation to the directives force, it will here be understood that this term refers to the whole external magnetic force actually effective upon the compass; this being known, in general, on shore, as the horizontal magnetic force of the earth, and on board ship as the resultant of the horizontal magnetic forces of the earth and ship, for the particular heading of the ship, the particular place, and the particular time under consideration.
And, similarly, in relation to the position of rest, it will also be understood that the zero-line of the card (supposed to coincide with the magnetic axis of the card), while at rest, is in the position of equilibrium in respect of all the forces, magnetic and otherwise, acting upon it; this being known, in general, on shore, as the magnetic meridian or direction of the earth's horizontal magnetic force, and on board ship as the deviated direction of the compass, or resultant direction of the directive force on board.
The moment of motive force, the directive force remaining the same, varies, therefore, with the magnet-power of the card multiplied into the sine of deflection. It is greatest as the card is deflected 90 degrees from its position of rest, when the moment of deflection is equal to the product of the magnet-powder by the directive force, less the moment of resistance; and it is least as the card comes to rest, at or near its previous position, when the moment of deflection is equal to zero; the moment of the motive force being reduced to an equality with the moment of resistance, whatever that may be.
Accordingly, we have for the sine of the angle of set (or terminal angle of deflection), or for the arc of set, if the angle be not very large, the moment of resistance divided by the product of the magnet-power and directive force.
In order, therefore, that the card may always return exactly, or very nearly, to its position of rest, whenever deflected from it, it is necessary either that the moment of resistance be extremely small in comparison with the product of the magnet-power by the directive force, or that this product be extremely large in comparison with the moment of resistance.
Now, the directive force of the earth, in different places traversed by the navigator, varies from about one-half to about twice its mean value; while the directive force on board, especially of iron-built ships, may vary quite as much on different courses of the ship, even in the same locality. Consequently, if, from developed defects of the compass, the moment of resistance be unavoidably large, or, on the other hand, if the directive force on board be much below its mean value, the angle of set, even with the magnet-power of the card unimpaired, will become so much the more appreciable.
It is, therefore, quite essential to the reliable behavior of the compass, under the varying circumstances of a ship's cruise in different parts of the world—
First, that the magnet-power of the compass-card should be as great, in every case, as can be conferred upon it, compatibly with other necessary conditions; or, in other words, that our aim should be to secure not only enough magnet-power for ordinary or average circumstances, but a surplus or reserve for extraordinary occasions of special requirement.
Secondly, that the magnet-power of the card should be as nearly permanent as can be realized through the formation of the card-magnets; and, to this end, that the greatest care should be used during every stage of that process.
Estimating the magnet-power of a compass-card.—These conditions will, perhaps, be more truly appreciated if we consider for a moment the means by which the magnet-power of a compass may be correctly estimated. Three different methods may be employed, with greater or less convenience, for this purpose.
First, the method of deflections.—By this (statical) method, the compass-card whose magnet-power is required is made to deflect a standard magnetic needle at a certain measured distance between their respective centers. The magnet-power of the card is then equal to one-half the cube of the distance multiplied by the product of two factors, of which one is the tangent of the observed deflection, and the other is the directive force; it being understood that the card is so presented toward the magnetic needle that its zero-line is in the magnetic equatorial through the centre of the needle."
Secondly, the method of oscillations.—By this (dynamical) method, the compass-card is made to oscillate in its own plane, and the time of one oscillation noted. The magnet-power is then equal to three-tenths of the moment of inertia of the compass-card, divided by the product of two factors, of which one is the square of the oscillation-time and the other is the directive force; it being understood that the units of distance and time are the foot and second.
For the same card, the moment of inertia is constant, and the magnet-power is proportional inversely to the square of the oscillation-time multiplied by the directive force.
For different cards, with the use of the same auxiliary weight, the moment of inertia of the card may be expressed in terms of the moment of inertia of the weight; and the magnet-power is then proportional inversely to the difference in the squares of the oscillation-times (with and without the weight) multiplied by the directive force.
Thirdly, the method of torsions.—By this (also a statical) method, the compass-card is suspended in a torsion-balance, and the moment of deflection at any angle balanced by the corresponding moment of torsion. In this case, the magnet-power of the card is equal to the moment of torsion divided by the product of two factors, of which one is the sine of the deflection and the other the directive force.
For the same conditions of torsion, the moment of torsion varies directly as the angle of torsion ; and, accordingly, with the same angle of deflection, the magnet-power of any compass-card is proportional directly to the angle of torsion divided by the directive force.
The directive force is thus seen to be an element in each of these methods, as, indeed, it must necessarily be in every estimate of the magnet-power of a compass-card, or of any simple magnetic needle. Still, so long as the required determinations are made at the same place on land, it will be sufficiently exact, within moderate periods, to regard the directive force as constant, in which case the proportionality of the magnet-power is independent of the directive force. But, as already mentioned, since this element, under the combined influence of geographical position and the ship's heading, may vary in a several-fold ratio, a proportional change must result in the remaining elements of the determination, provided the magnet-power of the card is unchanged. Hence, generally, in order to any reliable estimate of the magnet-power of a compass, under the varying circumstances of its use at sea, the directive force mud always he known as a necessary preliminary.
Now, there is no physical difficulty in obtaining absolute determinations of the directive force by well-known methods whenever required; and with this element absolute determinations of the magnet-power of a compass-card could be had, if desired, by either of the foregoing methods. Nevertheless, for all practical purposes connected with the use of the compass, it is always quite sufficient to obtain relative values of the magnet-power for the directive force taken as unity at some convenient initial point.
It would lead me too far from the immediate object of this communication to enter more into details, of a purely determinative kind, in relation to the magnet-power of a compass. It may be sufficient to say, in passing, that the relative directive force, either on board or on shore, may always be found by very simple means, and with sufficient precision for the purpose in view. With respect to the several methods indicated, the second has certain advantages for use at sea: first, that no auxiliary instrument is required; and, secondly, that the removal of the card from the compass-bowl is unnecessary, which, in the case of the liquid compass, is attended with some inconvenience.
Developing the magnet-power of a compass-card.—Now, with respect to the two conditions of the magnet-power previously noted, it will be evident, from the second of the preceding methods, that the question of gaining magnet-power in a compass-card will depend on the possibility of producing a greater increase in the moment of inertia of the card than in the square of its oscillation-time. If, by introducing a different weight and distribution of steel, the moment of inertia is thereby increased m times, while the square of the time is only increased n times, n being less than m, there is a gain of magnet-power in the ratio of m to n.
Practically, this question resolves itself into two parts: first, that of increasing the magnet-power in the formation of single magnets of given weights and dimensions; and, secondly, that of distributing the magnets upon the compass-card in a manner to increase the magnet-power of the card.
First, the development of magnet-power in the formation of single magnets.—This question, which is essentially one of experimental research, has been the subject of numerous special investigations; but by far the most exhaustive inquiry which has ever been made, although open, perhaps, to criticism on certain unimportant points, was that of the late Rev. Dr. Scoresby, from whose elaborate research the following conclusions may be summarized, as applicable to our present subject:
1. That the selection of steel for compass-magnets should be made from that known generally as the "very best," in the form of thin plates.
2. That the steel, after being cut into pieces of the requisite length and width, should be hardened uniformly throughout, and only annealed or tempered sufficiently to prevent too great brittleness.
3. That the hardened laminae should be magnetized to their utmost capacity by the most powerful inductive action at command, and each lamina separately tested for magnet-power.
4. That the magnetized laminae, after being laid together, in contact, with the like poles pointing in the same direction, should again be separately tested for magnet-power, and all rejected that show any sensible deterioration.
5. That the proved laminae should finally be built up in magnet-piles of two or more lamina in each; it having been conclusively shown that a compound magnet, consisting of several proved magnetized laminae, takes on a higher development of magnet-power than a simple magnet, in one piece of the same weight and dimensions.
It is not, however, to be understood that the gain in magnet-power from piling is proportional to the number of laminae in the pile; on the contrary, with equal increments of steel, the corresponding increments of magnet-power are successively smaller, decreasing, approximately at least, in a geometrical ratio with the number of laminae added to the pile; so that the practical limit of available gain in this manner is soon reached.
The conclusions of Scoresby, established more than thirty years ago, with respect to the formation of compass-magnets, have been frequently confirmed, although little has been added thereto since that time.
These several conclusions may, therefore, be adopted, until, at least, we are better informed, not only as the rules of procedure in obtaining compass-magnets of the highest intensity, but as also generally favorable to securing them of the greatest permanency.
At the present time, with the use of the comparatively unlimited resources of electro-magnetic induction, the means of magnetization are greatly in advance of those employed by Scoresby. Quite recently a means of heating and tempering the laminae for compass-magnets has been used by Mr. E.S. Ritchie, which ensures much greater uniformity, not only in the distribution of the degree of hardness sought, but also in the subsequent magnetization.
Secondly, the development of magnet-power in the distribution of magnets upon a compass-card.—If a magnet of uniform section be placed across the centre of a card-circle, its length being equal to the diameter of that circle, its magnet-power and weight will be proportional to the diameter, and its moment of inertia to the cube of the diameter. If, now, we conceive this magnet to be moved in either direction outward, parallel to its first position, taking up positions and reductions of length according to the successive chords, its weight and magnet-power will progressively decrease in proportion to the cosine of the angular distance of the chord from the diameter, while its moment of inertia will progressively increase in proportion to a certain function of that angle, reaching its maximum at an angle of 45°, after which it will diminish, till, at the angle of 80°, the chord and all that depend on it vanish together.
Thus, with a magnet at the angular distance of 45° from the centre, its weight and magnet-power are each decreased to 0.7 and its moment of inertia increased to 1.4 of their values, in comparison with a magnet equal to the diameter of the centre; and if two such magnets be placed upon two equal parallel chords at 45°, each of those quantities will be doubled, or their weight and magnet-power each be 1.4 and their moment of inertia 2.8. Hence, it may be concluded that, by placing magnets symmetrically on equal parallel chords, it is possible to gain in magnet-power, though at the expense of additional weight to be carried by the compass-card.
It will be shown hereafter that there are certain considerations which establish a choice of these symmetric chords for magnet positions. There are two such arrangements which are substantially equivalent, namely, the single pair on chords at 30°, and the double pair on chords at 15° and 45° respectively. The following table illustrates these several relations at one view:
Distribution of Magnets on a Card.
Designation |
One magnet at center. | One pair of magnets on chords at 30° | Two pairs of magnets | ||
On chords at 15° | On chords at 45° | Sum | |||
Magnet-power | 1.0 | 1.7 | 1.9 | 1.4 | 3.3 |
Moment of inertia | 1.0 | 2.6 | 2.2 | 2.8 | 5.0 |
Weight of magnets | 1.0 | 1.7 | 1.9 | 1.4 | 3.3 |
Thus it may be seen that, in assuming a certain practical limit to the increase of section by piling magnetic laminae in the formation of single magnets (which would be essentially the same for lengths varying between 1.0 and 0.5), and distributing these magnets of equal section upon the parallel chords, according the either of the above-named systems, there will be a material gain in magnet-power, and a larger gain in the moment of inertia, in comparison with the single magnet at the center.
II.—SENSIBILITY OF THE COMPASS.
If a compass-card, on being deflected to any extent in either direction from its position of rest, and then left to itself, return precisely to that position, it may be said to possess perfect sensibility; but, on the contrary, if it fail to come precisely to its previous position, the angle of set by which it deviates from that position may be called its defect of sensibility.
Now, if there were no resistances to the motion of a compass-card, and if it had any appreciable magnet-power, it would invariably return to its previous position of equilibrium, whenever deflected from it by virtue of the motive action of its moment of deflection, and, consequently, no defect of sensibility could arise.
But, in point of fact, it is a physical impossibility that there should be no resistance to the motion of any body within our immediate cognizance; and, consequently, we must expect, in accordance with our previous assumption, an angle of set, or defect of sensibility, whose value is represented by the moment of resistance divided by the product of the magnet-power of the card and the directive force acting upon it.
There are, in reality, two different resistances to the motion of the card: one is the friction of the pivot; the other is the resistance of the medium, air or liquid, in which the card moves within the compass bowl. The former is a constant; the latter is a variable, depending on the velocity of the card at any particular instant of its motion. The moment of resistance, already referred to, consists, therefore, of the moment of friction at the pivot and the moment of resistance in the medium, both moments being referred to the point of the pivot as the centre of moments. The moment of Resistance opposes the motion alike during the increase and during the decrease of the angle of deflection.
The moment of friction consists of three factors: the pressure between the rubbing surfaces, the mean radius of the area in contact, and the coefficient of friction; the latter depending on the physical qualities of the pivot and cap, such as hardness, smoothness, etc. All these factors are essentially constant for the same card, except as they may be liable to change with changes of condition.
The resistance of the medium is more complex; for it not only involves several distinct elements, but its law of action is somewhat uncertain under considerably varied circumstances of the form and velocity of the moving body. Nevertheless, it appears to be certain that the resistance of a medium, properly so-called, is solely a function of the velocity of the moving body, involving no absolute term independent of that velocity. As to the form of this function, it is far less certain; but we are justified, from the results of experimental research on this subject, in concluding that the moment of resistance of the medium to the motion of a body of unyielding form, like that of a compass-card, is represented by a product of five factors—the square of the velocity of the card, its section of resistance, the mean radius of that section, the density of the medium, and the coefficient of resistance; the latter depending on the form of the card, and possibly, also, on the velocity, in view of the considerable variation in this element during the motions of the card. Of these factors, all but the first and fifth are sensibly constant for the same compass; and of the fifth there is only some doubt whether it can always be expressed as a constant for the same compass, or must be modified somewhat for the variable velocity.
I have entered somewhat more into dynamical details, especially in regard of these resistances to the motion of a compass-card, than might at first sight seem to be necessary; but, in order to form an intelligent judgment of the conditions which should control the construction of tie marine compass, we must take into consideration the laws of the resistances to the card motion, and these cannot be duly appreciated without, at least, a definite recognition of all the elements of these resistances.
Now, with regard to the relation of these resistances to the sensibility of a compass, it will be evident, I think, that the resistance of the medium, however great it may be during certain stages of the motion of the card, cannot give rise to any part of the defect of sensibility: for, in the first place, in regarding this resistance as solely a direct function of the velocity, it must decrease with the velocity and completely vanish with the cessation of motion; secondly, and with still stronger reason, it should follow from the assumption that the resistance varies directly as the square of the velocity; that, as the velocity of the body diminishes in approaching its final position of rest, the resistance of the medium diminishes in the much more rapid ratio of the velocity squared; so that, with the last clement of the velocity, the last clement of this resistance is a quantity infinitely smaller in comparison.
Hence, it must be concluded that the resistance of the medium, whether air or liquid, has no influence whatever on the ultimate angle of set by which a compass-card deviates from a previous position of rest after being deflected from it.
It is otherwise with the friction at the pivot; for this being a constant force wholly independent of the velocity, it remains the same, during the smaller elements of the velocity, as during the most rapid motions, until it finally comes into equilibrium with the motive forces, and there results an angle of set, or defect of sensibility, which is represented by the moment of friction divided by the product of the magnet-power and directive force.
There may be a question whether, in addition to the friction of the cap upon the pivot, there may not be a certain amount of friction due to the action of the fluid in the cap immediately surrounding the pivot. This point is involved in some obscurity at present. It is not easy to separate a possible frictional resistance like this from what is recognized as the resistance of the medium proper; so that it is quite probable, whenever the former has an appreciable value, that it should be merged in the latter, in the results of experiment.
Consequently, the conditions most favorable to the sensibility of a compass appear to be these:
- That the pressure of the card upon the pivot and the area of the surface in actual contact between the cap and pivot shall both be as small as possible.
- That the material of the cap and pivot shall be as hard, as true to form, and as smoothly polished as possible; and,
- That the magnet-power of the card shall be as great and as permanent as possible.
The pressure upon the pivot remains unchanged for the same compass-card; but both the mean radius of the rubbing surface and the coefficient of friction are liable to increase—the first from the wear of the material and the second from the irregularities and roughness of the wear. The magnet-power is liable to decrease—in some cases very seriously from original defects in the formation of the card-magnets, and in others from accidental causes incident to the handling of the compass on board ship.
But whatever the angle may be which represents the defect of sensibility, either at the outset or as the result of subsequent changes in the compass-condition, it is always an error of the compass. Moreover, it must be regarded as one of the most dangerous errors to which the compass is liable; because, whenever the actual condition of the compass is unknown, its value is as uncertain as that of the function upon which it depends; and this may vary from an extremely small quantity, when the sensibility is practically perfect, to a quantity as large as unity, when the sensibility is nothing.
In passing from the present topic, a remark may be permitted on the preceding definition of sensibility. It will hardly escape notice that the definition were given is not strictly in accord with a prevalent habit of verbal expression, not only among nautical men, but in ordinary popular language. Thus, a compass is said to possess sensibility when "It is lively," when "it moves quickly," etc., without regard, so far as I am aware, to the condition which I have regarded as essential to the idea of compass-sensibility.
Now, a compass-card, when nicely balanced at a jeweled cap upon the point of a hardened pivot, is extremely susceptible to motive influences from purely mechanical causes, independently of any magnet-power whatsoever in the card. The slightest disturbance actually applied to it may be sufficient to set it in motion. It is true that the motion will be different in certain respects when the card at the same time possesses any magnet-power. But the mere excitability of a compass-card, however great, and whether resulting in vibratory or other motions, cannot be regarded as a true or sufficient criterion of its sensibility from a magnetic point of view.
The intrinsic property of a compass-card, alike with that of a simple magnetic needle, is its tendency to return to its position of magnetic equilibrium whenever deflected from that position; and this is realized under the combined influence of the exterior directive force and its own magnet-power. If it fail in any degree to do this, its most characteristic, not to say its most useful, property is so far imperfect. The question, as it seems to mc, is not whether the card is more or less excitable in its movements about its position of rest—for this may depend on several distinct circumstances—but solely whether, in whatever way done, it accomplishes unerringly, and with a nicety of precision that admits of no doubt, its prime function. When it does this—which it never can do except by accident, unless the resistances to motion are so small in comparison to the motive forces as to be uninfluential—then I think that the specific term sensibility is both significant and appropriate as the expression of such a fact.
III.—STEADINESS OF THE COMPASS.
A compass-card is said to be stable, or steady, when it maintains its position of equilibrium, under the magnetic forces which act upon it, without sensible disturbance by the various mechanical influences which are liable to be called into action on board ship. A compass may possess sufficient magnet-power and perfect sensibility, and yet be so deficient in steadiness, during the rolling, pitching, yawing of the ship, as to be practically useless. But any apparent compatibility of deficient steadiness with perfect sensibility can only exist for a very brief period ; for the effect of much motion of the card must be to blunt or otherwise injure the pivot, or to wear the cap, with the inevitable consequence of increasing the friction, and thereby diminishing the sensibility.
Card-unsteadiness is a mechanical difficulty, and the remedy must be mechanical, so far as it is practicable to have one without compromising the sensibility.
There are two conditions of steadiness, which, at the outset, are always applicable to the cards alike of dry and liquid compasses. These are—
First, that the card shall have a tendency, whenever tilted to one side, to return to its position of horizontal equilibrium. This condition is satisfied by placing the centre of suspension well above the centre of gravity of the card, and also, in the case of the liquid compass, above the centre of buoyancy in the card.
Secondly, that the card shall have no tendency to rock in one direction more than in another; that is to say, no tendency to wabble about any of its diameters. And this condition requires that the material of the card shall be so distributed as to give equal moments of inertia about all its diameters. It is satisfied by arranging the relatively heavy card-magnets in one or more symmetrical pairs, on equal parallel chords of the card, at certain calculated distances from the centre.
But these conditions, although on the side of stability, so far as they go, and quite essential to a well-made compass-card of any kind, fall far short in practice of realizing even tolerable steadiness in the air or dry compass, for, while the card may not be liable to wabble, it is still prone to rock in every direction, and, although prevented from actually tilting over, it is liable to spin entirely round in its own plane under the influence of sharp jars or shocks.
Various remedies have been proposed at different times for this serious defect of the air compass. One of these consisted in fixing several projecting pins upon the upper surface of the card, which, by their friction against the glass cover above, might subdue excessive whirling and rocking motions, although no actual contact need exist while the card was in its more quiet and normal condition.
Another compass, well known and still in use, of a celebrated maker, has the provision of a very heavy card, weighing not much less than ten ounces, supported on a fixed spindle passing through it, with upper and lower bearings; and this arrangement, whenever its resistance to motion proves insufficient, is aided by a friction-brake, which may be turned on ad libitum, until the card becomes quite steady, as it undoubtedly should with the means at command.
The simplest and probably the best provision of this kind is that of a merely heavy card, with enlarged bearing surfaces at the pivot.
These, however, are not a tithe of the different devices which have been resorted to from time to time, as remedies for the evils of an unsteady compass. I have referred to them merely as illustrations of a kind of relief much resorted to even by intelligent navigators of the present day, and yet they certainly appear no more rational than the recourse of the less-informed skipper, whose little craft, dancing like a cockle-shell upon the waves, infects his compass with an excitement which he endeavors to allay by putting brick-dust in its cap. In principle they are the same. The remedy, so far as it proves effective, consists in the production of a moment of friction capable of counteracting the mechanical excitements to motion.
Without entering into any descriptive details, I think it may be said that the prevailing idea of these provisions is that of a heavier card, with more powerful magnets and the use of more rounded pivots. But, with the increase of pressure and bearing-surface at the pivot, there comes a proportional increase in the moment of friction; and thus, while the magnet-power is increased in a certain ratio, the moment of friction is augmented in a much higher one; so that, on the whole, there results at the outset is considerable sacrifice of sensibility, attended by a corresponding error of the compass. And this is not all, for these heavier weights develop proportionally greater wear at the pivot; and, if to this be added the possible deterioration of the magnets, it is not difficult to see that, even if the defective sensibility be tolerated at the beginning, the error from this source is liable to become so great, and withal so uncertain, as to make the advantage gained in mere steadiness (never, I believe, very satisfactory at the best) of doubtful value, in view of the possibly very serious sacrifice in precision.
If practicable, therefore, such a remedy should be found for unsteadiness of the compass-card as shall not impair its sensibility. And this we have by combining with the two preceding conditions a third, namely, that of placing the compass-card in a liquid instead of a gaseous resisting medium.
By the use of a liquid rather than an air medium, we gain the advantage of the greatly-increased resistance due to the superior density of the former, which, for the liquid likely to be employed, would not be much less than 800 times that of the air. The law of the resistance would be the same in both.
With this provision, the more violent the impulse to motion, the more energetic the resistance; since, as the velocity of the card increases, the resistance of the medium increases in the more rapid ratio of the velocity squared, while, as already noticed, the resistance decreases in the same rapid ratio as the velocity becomes less.
This is illustrated by a glance at the two horizontal rows of figures, of which those in the upper row represent velocities, and those in the lower row the corresponding proportional resistances:
Velocities 32, 16, 8, 4, 2, 1, ½, ¼, 1/8, 1/16, 1/32
Resistances 1024, 256, 64, 16, 4, 1, ¼, 1/16, 1/64, 1/256, 1/1024
Thus, a velocity represented by 33 encounters more than 1,000 times the resistance that a velocity of 1 encounters; while a velocity diminished to 1/32 encounters a resistance of less than 1/1000 of that due to a velocity represented by 1.
Hence the admirable facility with which a liquid compass may adapt itself to such opposite requirements; in one case presenting the most effective resistance for the destruction of all actual motions of the card, however great; in the other, offering the least possible obstacle to such motions, when in small arcs about the position of rest, whether in their incipient or terminal stages.
Nevertheless, the advantages even of a liquid medium arc greatly enhanced by a certain auxiliary provision, of sufficient importance to be regarded as a fourth condition of steadiness, to wit, that of the use of a buoyant skeleton-card with a minimum pressure at the pivot.
With this provision, the resistance to circular motions is greatly increased, not only from the larger effective section of the card, but also from the larger coefficient due to its skeleton-form; and, at the same time, the evil effects of the severe vertical shocks upon the pivot, experienced by the heavy disk-like cards, are greatly mitigated, in a higher proportion even than the reduction of pressure at the pivot.
Of course, the well-known advantage of the gimbal action is not to he overlooked as a fundamental condition of steadiness. But this provision is a condition of all marine compasses, besides being entirely outside of the compass-bowl. Without this, or its equivalent, no other provision of compass-steadiness would be of any avail whatever, and even the existence of such an instrument as the marine compass impracticable. I have thus presented an outline at least of the considerations which, in my judgment, should control the construction of the marine compass, upon the basis of the three fundamental properties assumed at the outset to be essential to its most perfect action on board ship.
IV.—THE NAVY COMPASS IN THE LIGHT OF THE PRECEDING REQUIREMENTS.
The Navy compass, as already intimated, has the distinctive peculiarities of a buoyant card in a liquid resisting medium ; the mean density of the card being so adjusted to the density of the liquid as to produce a small downward pressure upon the pivot in the ordinary forms of ship and boat compasses, or a small upward pressure against the pivot in the special form of "tell-tale" or cabin compass. The compass-bowl is provided with a self-adjusting expansion-chamber, by means of which the bowl is kept constantly full, without the show of air-bubbles on the one hand or the development of undue pressure on the other, from changes of temperature.
The ship compass of general use has a 7 ½-inch skeleton-card, with provision for one symmetrical pair of magnets, a division on the outer ring to quarter-points, and a card-circle adjusted to the ring, which is divided to half-degrees. The bowl-circle, or outer edge of the rim upon the bowl, is made rigid and turned strictly 1o gauge, so as to admit of the interchange, from one bowl to another, of every azimuth-circle of its class. The compass is alike used in the steering-binnacle or for azimuth purposes.
I shall now briefly consider, under the three general heads previously treated, how nearly this compass appears to be capable of satisfying the conditions therein set forth.
First, with respect to magnet-power.—The magnet-system of this compass consists of two equal compound magnets, enclosed in parallel tubes in the two chords of the circle, a little within the angle of 30 degrees from the parallel diameter. Each magnet is built up of six laminae; each lamina being 6 ½ inches long, 7/16 of an inch wide, and about 1/40 of an inch thick. Each compound magnet weighs about 880 grains, or a little less than two ounces, with but slight variations.
The steel of which these magnets are made is that known in commerce as "Stubb's sheet," which, from numerous experiments by Mr. E.S. Ritchie, has proved to be the best for this purpose, not only for its uniform excellence, but for its magnetic capacity in both intensity and permanence. In this Mr. Ritchie has but confirmed the conclusions of Doctor Scoresby, of thirty or more years ago, as to the superior qualities of this (English) steel for magnetic purposes.
The laminae, having been cut to the proper size, are hardened and tempered throughout their entire length, the process being so conducted as to secure a remarkable degree of uniformity in the results. The magnetization is then effected by means of a very powerful electro-magnet to their utmost capacity. After this, the laminae are separately tested for their relative magnet-power by a deflection-needle, and the angle of deflection marked on each; and, finally, they are laid aside for a little time in promiscuous contact. As required in the formation of card-magnets, these laminae are next subjected to a careful scrutiny, being taken, one by one, and again tested for magnet-power; and every piece which shows any sensible falling off, as compared with the previous test, is thrown out.
Although I was hardly satisfied, a year ago, with certain details in the formation of our card-magnets, I am convinced that the present process, as just described, is substantially in accord with our best knowledge on this subject, and is destined to leave little to be desired in point of completeness and thoroughness for the end to be attained, namely, to secure the most powerful magnets, compatibly with the condition of the greatest permanency, for given weights of steel.
As to the actual magnet-power of the Navy compass, it was important to know how it compared with that of other well-known compasses. For the purpose of a comparison, I selected two 7 ½-inch cards of well-known English makers, the best of their kind, and designed especially for steadiness, as "heavy cards." One is designated as card "B, 468," the other as card "D, 305," the latter being a spindle card for double bearings. The former has two magnets, and the latter four, in symmetric pairs. Both cards belong to compasses of the collected specimens in my rooms at the Bureau of Navigation, and both appear to be in good condition; but whether either has suffered any loss in magnet-power, as compared with its original condition, I am of course unable to determine. The Navy card is one of recent make.
The tensions for this method were reduced to a condition of approximate equality by the application of weights.
The results by the three methods are in quite close accord, with the exception of that for the card “B, 468,” by the method of torsions, the somewhat smaller value of which being due probably to the less favorable conditions under which the observations by torsion were made for that card.
It will thus be seen that the magnet-power of the navy-card, while somewhat more than twice that of card “B, 468,” is less than that of card “D, 305,” in the ratio of 1,000 to 1,450.
It will also be noticed, in comparing the oscillation-times and moments of inertia in the second and fourth columns of Table II, that a longer oscillation-time is not a certain indication of a lower magnet-power, unless due account be taken of the moment of inertia. Thus, the oscillation-time of the navy-card is about thirteen per cent greater than that of card “B, 468,” but its moment of inertia is nearly three times as great as that of the latter. On the other hand, as compared with the card “D, 305,” the oscillation-time of the navy-card is still greater, buts its moment of inertia is actually smaller, so that in this case the magnet-power is smaller, as it should be, than that of the card “D, 305.”
Although, as shown by these observations, there is a good comparative degree of magnet-power in the navy-card, my principal doubt is, at present, in regard to the question whether we have yet reached the practical limit of magnetic development to which the card may be judiciously pushed. No case even of apparent deficiency in the magnet-power of this card has ever been brought to my notice; and my doubt on the point is, therefore, not based on the supposition of actual deficiency in this respect for ordinary circumstances, so much as on the conviction, heretofore expressed, that the compass-card should always possess a liberal reserve of magnet-power—up to the very limit which may be imposed by other conditions of card-construction—in order to provide for those large fluctuations in the directive force the effect of which might be to seriously diminish the moment of motive force, and thus to proportionally increase the defect of sensibility.
With respect to the question of magnetic permanency, our experience is too recent with the present process of magnet-formation to permit the expression of any opinion as based upon actual results. This, however may, I think, be said: that the recent changes in some of the details of that process are precisely such as, while obtaining a somewhat higher average of magnet-power in the magnet-piles, are well adapted to secure the most reliable state of permanency.
It should not be inferred that our previous experience has been particularly unfavorable with the results of the old process. So far as I am aware, not an instance has occurred, within several years past, of a reported discovery of any serious declension in magnet-power of the navy compass. Still, this is a kind of negative evidence, to which I am inclined to attach very little value, in the face of one positive fact to the contrary; and I hope to have the means hereafter of ascertaining the facts, on the return of our compasses to store after considerable periods of service on board ship.
In this respect, as in many others, there is an important advantage in favor of the navy compass—that the compass-card, being always delicately balanced on its pivot in the bowl, is in the best practicable condition for maintaining its magnet-power, other things remaining the same.
But, whatever may be revealed hereafter by a closer enquiry into the facts of our navy experience with the present form of compass, I am fully convinced that magnet deterioration is a much more prevalent and more serious evil than it is generally supposed to be by nautical men. I have had occasion, within a year past, to notice a number of instances of this kind, some of which were serious enough, and in one or two instances where I least suspected it, and by which I was considerably astonished.
The whole subject of the magnet-power of a marine compass, in its twofold aspect of intensity and permanency, has appeared to me of such fundamental importance that I have determined to devote some time to its special study, with the hope that I may be able to clear up certain points not now as well established as I should be glad to have them.
Secondly, with respect to sensibility.—It will not be very difficult to understand why the navy compass should be expected to possess a high degree of sensibility.
Keeping in mind the condition already stated, that the defect of sensibility is equal to the moment of friction divided by the product of the magnet-power and directive force, let us consider the actual relations of these elements in the navy compass.
Now, as to this compass, the mean density of the submerged card admits of being so adjusted to the density of the liquid as to secure any desired buoyancy, and consequently produce any desired pressure of the card upon the pivot, however small, and whether upward or downward.
The minimum pressure at the pivot of the seven and a half inch card has thus been adjusted to about sixty grains, at the mean temperature of sixty degrees Fahrenheit, in order to provide for the variations of temperature and consequent changes in the density of the liquid ordinarily encountered at sea. It is necessary and sufficient that the least pressure to which it may ever be reduced shall be such as to secure actual contact at all times between the cap and the pivot; and, on the other hand, it is desirable that no greater excess of pressure should be had, beyond the prescribed mean limit, than what is actually sufficient to satisfy the first condition.
It should be understood that these conditions of the card-pressure at the pivot are alike applicable, or nearly so, to the ordinary case of the downward pressure and to the special case of an upward pressure.
The relations of these pressures, downward and upward, to certain specified temperatures, for a liquid of normal mixture, at a pressure of fifty-eight grains at sixty degrees, have been noted by Mr. Ritchie, as shown in the subjoined table:
Temperature of liquid | Pressure at pivot | |
| Downward | Upward |
Deg. Fahr. | Grains | Grains |
85 | 88 | 28 |
60 | 58 | 58 |
20 | 27 | 89 |
13 | 18 | 98 |
Again, so far as the choice of materials for the cap and pivot and the forming of the bearing-surfaces are concerned, the advantage is still with the Navy compass; for, inasmuch as the bearing-pressure of the card is so greatly reduced, it will be allowable to use still harder materials and more sharply defined pivots than would be admissible in air compasses of the same size, whose lightest cards seldom fall below fifteen hundred grains; and hence it follows that not only the mean radius of the bearing-surface, but the coefficient of friction, may be reduced to smaller values than they could have with the best possible form of air-compass card.
Accordingly, the moment of friction of the Navy compass is materially smaller than that of any air compass. Thus, without placing any estimate on the possible reduction of the two elements just named, the pressure alone, as compared with that of the lightest air-compass card, is not more than one twenty-fifth part, while it may be less than one-sixtieth part as compared with that of the heavier cards.
And to this must be added the further advantage in favor of the Navy compass: that, in consequence of the extremely small working-pressure of the card, the wear of the cap and pivot is so small, even during all the vicissitudes of the longest cruise, as not to materially increase the friction or diminish the sensibility. In some instances a perceptible wear of the agate in the cap has been observed on the return of the compasses for examination; but in general the change is scarcely appreciable.
We have, then, in brief, two signal advantages of the Navy compass in point of sensibility: first, that of the extreme smallness of the moment of friction; and, secondly, that of the proportionally small liability to change of that friction. And the second is scarcely inferior in importance to the first.
How much should be added, if anything, for the friction of the liquid in the cap, is a question which cannot be readily answered with the present state of our knowledge on this subject. That it must be very small, if, indeed, it be an appreciable element in the resistance of friction, appears quite certain, in view of such direct observations as I have been able to make on the sensibility of these compasses.
In order to illustrate the preceding view by the facts of experience, I shall first give the results of some recent inspection-tests for sensibility of a number of new Navy compasses received from the makers.
The test for sensibility consists in bringing the vertical cross-hair of a telescope into precise coincidence with a division on the card-circle—as, for example, one of the zero-divisions; then deflecting the card a few degrees to one side by means of a small magnet, and allowing it to come to rest, to note the angle of set or defect of sensibility. The card-divisions are half-degrees; and it is not difficult by means of the telescope to estimate tenths of a division, or twentieths of a degree, and to appreciate still smaller parts—as small even as one-sixtieth of a degree.
The tests were actually made by deflecting the card 3", first to one side and then to the other, waiting in each case for the card to complete its vibrations and come to a perfect rest, before noting the deviation of the zero-division from the cross-hair.
It is so seldom that any appreciable defect is observed in the tests for sensibility of these compasses that I am led to regard them, in their normal condition, as in this respect practically perfect.
I think it needs but a single observation, with the cross-hair of a telescope nicely adjusted upon a division on one of these cards, to be convinced of its exceeding delicacy of action. By observing in this manner the behavior of a card after being deflected, as it approaches its final position of equilibrium, it will be seen to perform a series of minute oscillations about that position, so small and relatively so slow as to be scarcely appreciable by the unaided eye; suggesting most conclusively, I think, as already indicated on theoretical grounds, not only that the resistance of the medium at this stage is practically evanescent, but that the friction itself must be extremely small, in order that the moment of the motive forces acting at such small angles should be capable of overcoming it.
It might naturally be asked, after what has been said of the Navy compass. How does the test for sensibility result when applied to other compasses? In answer to such a question, I present the results of a few observations upon compasses of different makers, from the collection at the Bureau. With the exception of the three Navy compasses, they are all imported specimens of English makers; partly liquid and partly air compasses. None has ever been in service, and all are kept -with the cards freely suspended upon their pivots.
In making the experiments the zero or N point of each card was first adjusted to nice coincidence with the east side of the lubber-line, and the defections were made with the aid of a small magnet.
Those experiments which are preceded by a * were made after having readjusted the N point of the lubber-line.
It has not been my purpose, in giving these results, to suggest comparisons which should be regarded as in the least degree invidious; nothing could be further from my own taste or the temper of mind with which these or any similar inquiries should be conducted. The compasses are of excellent workmanship, as those which I have seen of the well-known London makers generally are; and their deficiencies in this respect are to be attributed to the inherent defects of construction, if, at least, the preceding views are accepted; namely, to the sensibly large moment of friction (as compared with that of the Navy compass), and in one or two instances to the added defect of insufficient magnet-power, both of which, as we have seen, concurring in the production of an angle of set, or defect of sensibility.
Thirdly, with respect to steadiness—The Navy compass is hardly less remarkable for steadiness than for sensibility. For this there are several reasons:
First. In the elevation of its Centre of card-suspension, decidedly above both the centre of gravity and centre of buoyancy of the card.
Secondly. In the distribution of its heavy weights, not concentric with the centre, in two equal parallel chords, a little within the angle of 30 degrees from the parallel diameter, thus securing nearly equal moments of inertia about all diameters of the card.
Thirdly. In the use of a liquid-resisting medium, with the advantages resulting therefrom.
Fourthly. In the use of a buoyant skeleton-card, adjusted to a very small pressure at the pivot, from which result the several advantages in favor of steadiness already enumerated under the general head.
Fifthly. In the preponderating inertia of the liquid mass over its friction against the interior surface of the bowl, in consequence of which any sudden impulse given to the latter causes it to slip ever or round the liquid without communicating any sensible motion to it and through it to the card.
A remarkable illustration of the steadiness of the Navy compass came under my observation, a year or two ago, at the Messrs. Ritchie's, in Brookline. One of the earliest appliances devised by Mr. Ritchie, Senior, for the practical study of the behavior of a marine compass, is a very effective arrangement for testing its steadiness. This apparatus, which was erected in the attic story of the workshop, consisted of a strong frame-work, with moving parts on opposite trunnions, so as to admit of giving to a projecting head-piece rolling and pitching motions, mingled with occasional severe jars and shocks of the most exaggerated kind. He called it his "model of a ship"; but a ship could hardly live in a sea that would cause such motions.
I mounted this arrangement on one occasion with Mr. Ritchie, when we had one of the Navy 7 ½-inch compasses and one of the 7 ½-inch air compasses of the best construction. The two compasses were placed side by side on the projecting head-piece, about three feet apart. The effect, as seen by us from our more elevated position, was sufficiently striking. The card of the air compass not only would roll and vibrate in the most extraordinary manner, but frequently spin round and round; while the card of the liquid-compass had hardly any appreciable motion, the only apparent motion being a slight swing from left to right, and from right to left, and even this was synchronous with the alternate motions in azimuth of the head-piece on which the compasses were placed. Although I made no measurements to strictly confirm this impression, I could hardly resist the conviction that the small apparent motion of Die card was in reality due to the actual swing in azimuth of the lubber-line. It seemed to me, from such a test, that the steadiness of the liquid-compass might justly be regarded as sensibly perfect.
Before concluding my review of the Navy compass, in which I have not hesitated to set forth with some prominence its manifold advantages, I should not omit, I think, to mention its defect, not as a peculiarity of this compass, but as inherent to the construction of all liquid compasses. It is the practical difficulty in effecting readjustments of the card equilibrium, if found necessary for the correction of defective horizontality, resulting from any considerable changes in the magnetic dip.
This difficulty is simply one of inconvenience in opening the bowl to gain access to the card. It is easily enough managed by a person accustomed to it, with the appliances of the workshop, but it is a rather troublesome operation under different circumstances.
At present, the only remedy is the provision previously mentioned as one of the mechanical conditions of a steady card, namely, that of elevating the centre of suspension "well above" the centre of gravity of the card. By this means, it is intended to give to the card such an excess of stability as to overcome its tendency to obey the varying vertical component of the earth's magnetic force in different magnetic latitudes. That this provision is sufficient, within moderate limits of the change of dip, to prevent any appreciable error from defective horizontality, is, I think, quite probable; but how far it may be relied on, under more extreme changes, is a question that must be settled by careful observations with the opportunities that may be furnished by practical experience. So far as I am aware, not an instance has been reported to the Bureau of Navigation, during all the Navy experience with this compass of any difficulty in this particular, or of any apprehended error from this source.
But, as remarked in another case, merely negative evidence (or, in this case, the absence of any express reference to this matter in the reports of navigating officers) can hardly be accepted as conclusive that it is entirely safe to neglect this possible source of error when sailing in high southern latitudes.
Nevertheless, I apprehend no serious difficulty in providing a practical remedy for this trouble, should it ever be deemed necessary or expedient.
Another objection has sometimes been made to this compass, that it is inconvenient to handle, as a portable instrument, on a tripod for observations on shore. This I shall dispose of in a word by saying that I can conceive of no occasion for the use of any marine compass on shore, when a good surveying compass, costing less than a third as much, would not be greatly preferable for convenience in handling, facility of use, and precision of results.
CONCLUSION.
I had originally intended to include in this communication some remarks on the several instrumental errors to which the marine compass is liable, besides the defective sensibility already noticed—errors essentially of compass-adjustment; and, especially, to have given some account of the adjustments of the Navy compass and of the degree of precision actually attained, as shown by our recent inspection-tests; but this must be deferred to some other occasion. It may suffice to say that I believe the Navy compass is susceptible of a high degree of precision, and that it may be furnished to the service in a condition which shall be practically perfect in this respect. I have said nothing of the azimuth-circle, because the compass itself is what claims our first attention; it being wholly fallacious to expect reliable results with the use of an azimuth-circle, however excellent, upon a compass which is liable to serious errors of adjustment and of defective sensibility.
It has doubtless been presumed, from the general tenor of what has been said on compass sensibility, that considerable importance is attached to this as one of the instrumental errors of the compass. In reality, I believe its importance can scarcely be overestimated. The errors of adjustment, even when quite large, are at least of a fixed character, and, if once definitely ascertained, may either be disregarded in ordinary cases of setting courses and in working up, or possibly allowed for in cases of greater urgency. It is otherwise with the error from defective sensibility. This, even at the outset, may be sufficiently serious; but, whether more or less so, there can be no certainty with air compasses, however excellent in workmanship, as to the amount of this error a few weeks later after a little rough weather at sea.
It is sometimes asked, Of what use is all this refinement of an instrument (generally concluded to be incapable of precision), so long as the navigator is unable to profit by it, and when, if he could, he is well enough satisfied if he can steer his courses to the nearest quarter of a point?
To this it may be said that, if we concede the sufficiency of such steering in an open sea (although there may be some who might regard it as hardly close enough in these days of "swiftest transit by the shortest route"), how is the navigator to be certain of doing even that with a compass whose instrumental errors, unknown alike in name and amount, may be much greater than the assigned limit to his error of observation; and especially when, in addition to the assumed errors of observation and the unknown errors of the instrument, the compass error is further complicated by the uncertainties of the variation and the deviation?
My own conviction is, as the result of considerable study of the subject, that, in view of the inevitable errors of observation to which the compass is liable under the trying circumstances of its use at sea, it should be our object, in the first place, to insure in the construction of this instrument not only its practically perfect condition when put on board ship, but its continuance sensibly in that condition during at least one cruise of the ship; and, in the second place, to facilitate the determination of the magnetic variation and compass deviation, considered as compass errors in the reduction to the meridian, and to bring the uncertainties of these determinations within such definite limits as it may be possible to assign with a sufficient knowledge of the circumstances of the case.