644.—Where instruments are taken abroad mercury silvering may become spotted, so that a small store of mercury and tin-foil should be taken out with the sextant for resilvering. But it should be particularly observed that the mercury should never be placed in the same case with the instrument, as the smallest particle, if it touch the frame, will eat into the brass and destroy its strength. Sealing-wax dissolved in spirit answers for a varnish at the back of the foil fairly well after resilvering if proper varnish be not at hand. It is advisable before attempting to silver a sextant mirror to practise on a few slips of ordinary glass in order to get into the way of doing it. In modern practice base silver is deposited, and no mercury is used, but the process requires special skill.
645.—Adjustment of the Index Glass.—Hold the sextant clamped to about 60° in a horizontal position with the index glass near the eye. Look nearly along the plane of the glass in such a manner as to be able to see one part of the plane of the arc by direct vision, and another part by reflection of it at the same time. If the direct view and the reflected join in one line, and the arc appears as the continuity of a single plane, the index glass is perpendicular to the plane of the sextant. If this be not the case it can be adjusted by turning the set screw placed at the back of its upper centre, [Fig. 286] E, very gently.
646.—Adjustment of the Horizon Glass to Perpendicularity.—Place the vernier at zero. Hold the plane of the sextant parallel to the horizon and observe if the image of the horizon seen by reflection at the edge of the silver line coincides exactly with the image received directly through the plain part of the glass. If it does so the horizon glass is perpendicular to the plane of the instrument, that is, assuming the index glass is also perpendicular. In this adjustment it is well to rock the plane of the instrument say 20°, to see that the horizon is cut as a clear line about its horizontal position for this amount of angle. If the mirror be not perpendicular adjust gently by the single screw at the top of the horizon glass frame. If the horizon be not water, the sharp outline of any distinct distant object will answer, or a piece of fine string placed at a distance and stretched straight.
647.—Adjustment for Index.—This is the adjustment for parallelism of the two mirrors at the zero of the arc. The sextant is clamped at zero as before: the arc of the instrument is turned in a vertical position and the horizon again observed. If this appears to cut a clear line through the plain glass and the mirror there is no index error, and the planes of the glasses are truly parallel to each other in this position. If the line is not continuous adjust gently by the lower screw, [Fig. 288], at G.
648.—Adjustment of the Horizon Glass by the Sun.—This is a better adjustment than that given above, except that it introduces any error that may be due to the imperfection of the shades; and it is more difficult particularly for the first approximate adjustment. Arrange the telescope and shades so that a clear outline of the sun's limb may be observed without distressing the eye. Place the vernier at zero. Observe the sun, which will be most conveniently sighted at about 40° elevation, first with the plane of the frame vertical, and then horizontally perpendicular to this. If the sun presents a round disc in both these positions the sextant is in adjustment. If in the vertical position there appears to be a small notch at top and bottom of the sun's limb, the glass is not perpendicular to the plane of the instrument, and this requires adjustment by the screw at [Fig. 288] K. If notches appear at the sides of the limb when it is held horizontally there is an index error, which may be adjusted at G if it be small.
649.—Index Error after Adjustment Allowance.—The limb of the sextant is graduated 5° beyond the zero position when the glasses are parallel to each other. This is called the arc of excess. The vernier is also divided three lines beyond its zero position, which is marked by an arrow. These extra divisions are placed on the instrument for correcting the index error by measurement of the angle subtended by the diameter of the sun's disc alternately on one side and the other of the zero line, in which observations, if the two readings agree, the sextant must be in perfect adjustment; when they do not agree half the error may be adjusted by the horizon glass. The same observations may also be made with a bright star by setting the index alternately on one and the other side of zero. When the sun is used the reflected and direct images are brought together, so that the two suns that appear in the instrument just touch limb to limb, first upon direct reading and then upon the arc of excess. When the division is adjusted very nearly, any small error, plus or minus, may be allowed as a constant for all readings. In observations of the sun care should be taken that the eye is protected, both by the sun-glass cover to the telescope and by sufficient use of the shades.
650.—Adjustment of the Telescope to Set its Axis Parallel to the Plane of the Sextant.—In fixing up the instrument after manufacture, the ring standard which carries the telescope is set at a measured distance from the plane of the frame, so that the centre of the ring coincides with the height of the silver line cut on the horizon glass. This is necessarily a primary adjustment. For final adjustment the long, inverting telescope is screwed home in the ring, and the eye-piece which has two parallel wires across its diaphragm placed in it. The telescope is brought to focus on any distant object, the eye-piece being turned at the same time to bring the wires parallel with the face of the instrument. Two objects are taken subtending an angle of 90° or over,—as the sun and moon, or the moon and a bright star,—and the index is moved so as to bring the objects, say the limbs of sun and moon, in contact with the wire nearest to the sextant, and fixed there. Then by changing the position of the instrument a little, the images are made to appear upon the wire furthest from the sextant. If the limbs of the sun and moon still remain in exact contact as they appeared before, the axis of the telescope is truly adjusted. If the limbs of the two objects appear to separate at the wire furthest from the sextant, the ring-adjusting screw furthest must be loosened a very little and the screw nearest the sextant tightened the same amount. If the reverse, and the images appear to overlap, adjust in the reverse direction. By repeating this operation a few times the contact will appear to be the same at both wires, and the axis of the telescope will be in collimation, that is parallel with the plane of the instrument. After the telescope is truly adjusted it may be raised or lowered a little to make the reflected and direct images appear equally clear.
651.—Final Examination of the Sextant.—It will be readily seen that an instrument, although correct in theory but depending upon perfection of workmanship in centring, division, surface and parallelism of glasses, and also in its adjustments, can scarcely be brought to perfection. The errors generally increase from the zero point, where adjustments are possible, and are greatest at 140°. In the ordinary commercial sextant of the dealers the errors of centring alone are commonly 3 minutes to 5 minutes, with like errors in other parts. It is therefore better, where the sextant has to be absolutely relied upon, to subject it to actual trial. The zero point can be readily fixed by rules already given; besides this the meridian altitude of several bright stars subtending angles of about 30°, 60°, 90°, and 120° should be measured either from a clear horizon or from a mercury artificial horizon, to be described presently, for angles under 60°, and the errors plus or minus tabulated. The data for the meridian altitudes of certain stars upon any night may be taken from the Nautical Almanac, which will require correction for the latitude and longitude of the observer. This subject is too complicated to be entered upon in detail here. At the present time the National Physical Laboratory undertakes the examination of sextants for a moderate fee. This is effected by means of fixed collimators, [art. 229]. For angles distributed over the arc the parallax error is eliminated by placing the collimators in pairs. The N.P.L. certificate may now be had with good instruments when purchased. It may be noted that an originally well-made instrument retains its qualities for all time, the wear of such instruments being inappreciable.