edge of the frame and its shelf: when the side of the box containing the horizontal needle is aligned with the edge of the shelf, the deflection of the needle from the magnetic meridian becomes known, that is, the deviation at that point; a long ruler attached to the side of the box will conduce to accuracy of alignment.

266. Exploring the magnetic field around the compass. -The observations to be taken at each point are: first, the deviation; second, the time in which the horizontal needle makes ten (10) oscillations (and in this connection see Art. 230, and Tables 36 and 37); third, the Dip-that is, the maximum angle the direction the dipping needle will make with the horizontal plane, noting at the same time whether the needle swings in the vertical plane through the magnetic meridian or outside it, and if the latter, by what angle (to be read on the horizontal graduation of the dip circle); and fourth, the time in which the dipping needle makes ten (10) oscillations through a moderately small arc in the vertical plane at right angles to the one in which the angle of dip was observed said plane to be found by turning the dip circle through 90° of the horizontal circle from the position it occupied when indicating the dip.

The observations for dip, and for horizontal and vertical intensity, are to be repeated on shore in a spot entirely free from magnetic masses: such a spot can be found by taking two sets of reciprocal bearings on lines at right angles to each other with an azimuth circle and 7-inch compass; and if the lines are about 100 feet long, it should not suffice to take the reciprocal bearings at their extremities alone, but also at every 20 feet of their length. The Variation should be determined in this spot by a Timeazimuth of the Sun; and the magnetic heading of the ship might be determined by a like observation on board.

The theory upon which the oscillations of a needle.

indicate the intensity of a magnetic force is explained in Parts First and Second. In the case under consideration, it is not the absolute force of either Ship or Earth that is required, but the relative value of the former to the latter; therefore it is optional to assign any standard for the Earth: it is customary to call it unity. The magnetic forces aboard and ashore being to each other as the inverse squares of the periods of oscillation of the same needle in both places, we then obtain that of the ship in terms of the Earth's by assuming the latter unity, and making the proportion between the periods of oscillation on ship and shore for the horizontal needle, and also for the vertical needle. A comparison of all the ship's observations with those ashore will give the values of the Deviation, Dip, Horizontal and Vertical Forces of the Ship for the particular heading she had in dock: in the same locality, these will vary with every new heading; and it would afford a better knowledge of the matter if all the observations could be repeated with the ship in the diametrically opposite direction—that is, docked stern first, instead of head first in the same place. Indeed, like swinging for deviations, the more points upon which the observations are made, the fuller the information afforded; but, for large ships, to make it on even the cardinal points would involve so much labor and variety of docking opportunities, that it seems almost hopeless to ever attain such completeness. The partial examination on one heading, however, reveals the fact whether it is the large magnetic field of the hull alone we have to deal with, or the concentrated but often powerful pole of some individual mass of iron.

Other information, also, is derived from observations on one heading: 1st, if begun while the ship is on the stocks and continued at intervals until completed, they exhibit the gradual forming of her magnetic character and its variations; 2d, if conducted in different parts of the ship

while in dry-dock, they indicate which is the best magnetic place for the compasses; 3d, when made on the site of the standard compass, they afford the means of obtaining a Table of Deviations; and 4th, they enable the navigator to compensate that compass with magnets.

Results under the third and fourth headings are only rough approximations; but still cases may arise when even such are useful: they will be explained in Part Fourth, and should be employed only when accurate methods are unavailable; and when used, they should not be depended on longer than the first opportunity to replace them by reliable work.

In securing the binnacles to the deck, it is important that they be placed so that the keel-line of the compass shall be exactly in the fore-and-aft mid-ship line: the trace of the central vertical plane through the keel is generally marked by the Constructor on the hatchways while building; if not, he has data and appliances for readily doing it; and with this to work from, it needs only a theodolite, or compass, azimuth-circle, and tripod, and the knowledge possessed by any officer who has to deal with the matter, to set the binnacles properly in place.

In order, however, to indicate a mode of procedure, the following is quoted from an officer who has had much experience in the matter-Commander Diehl, U. S. Navy:

"The hatches are brought up from the keel in plumb lines during construction of the ship; determine middle points of the hatches on the upper deck, and place a vertical staff on each point; run a line through all the middle points; at selected positions on this line erect straight-edges perpendicular to the keel, and by means of these place the binnacles on the mid-ship line so that this shall divide them symmetrically and be traced as a chalk mark on the upper rim or surface of each binnacle-a guide for further work. Level each binnacle by means of a spirit-level.

"Lower the heeling-magnet well down into its tube, place the compass in the binnacle, and proceed to center as follows: Raise the heeling-magnet in its tube, and if any deflection of the card results, move the compass to one side or the other by means of its centering screws, until no movement or deflection of the card occurs while the heeling-magnet is raised and lowered; the compass is then centered; remove the heeling-magnet and cause the pinnacle to assume the same inclination as the deck.

"Place an azimuth-circle on the compass, and sight successively on the straight-edges erected forward and abaft the binnacle: the readings should coincide with the keellines of the compass, and should differ 180°; if not, the binnacle should be moved until this is attained."

CHAPTER XVIII.

PHYSICAL REPRESENTATION OF THE THEORY OF THE DEVIATIONS.

267. The magnetic make-up of a ship analogous to that of a steel bar.—A theory of magnetism was stated in Arts. 191 and 192: briefly, it was, that a steel magnet is composed of molecules, either permanently magnetized or girded by electric circuits either conception explains observed facts; that the alignment of the molecular axes develops the magnetic condition, and their heterogeneous mixture the neutral state. By analogy we may consider the steel ship a magnet whose distinctive character is not confined to the hull alone, but depends also upon a thousand individual masses of iron distributed throughout the structure: engines shafts smoke-stacks-turrets-conning-towers guns-masts-hoisting cranes-hatch coamings-beams-ventilators- stanchions:- these, and innumerable minor masses, are so many separate magnets whose efforts variously affect the general result.

As with a bundle of bar-magnets-when placed with like-poles together, they exert a certain force; but if one, two, or more be reversed, the effect is thereby weakenedso with the iron masses in a ship: some contribute to the prevalent magnetism, others contravene it.

Provided it is not within the concentrated field of one of these masses that we place the compass, we may consider their individualities merged in the two grand regions of magnetic polarity that pervade the ship; and thus