through it. The back horizon glass H is silvered at both ends; in the middle is a transparent slit, through which the horizon may be seen. These two glasses are set in brass frames, similar to that of the index glass, and fixed on movable bases, which are adjusted by screws so as to set the glasses in their true positions. In general there are three dark glasses or screens, I; two red ones, of different shades, and one green. Each is set in a brass frame, which turns on a centre, that they may be used separately or together. They serve to defend the eye from the rays of the sun during an observation. The green glass is peculiarly adapted to take off the glare of the moon, but may be used for the sun when much obscured by clouds. When these glasses are used for a fore observation, they are to be fixed as in figure 1; but when used for a back observation, they are to be placed at O.

The sight vanes, K and L, are pieces of brass, standing perpendicular to the plane of the instrument. The vane K is called the fore sight vane, and L the back sight vane. There are two holes in the fore sight vane, the lower of which and the upper edge of the silvered part of the fore horizon glass are equidistant from the plane of the instrument, and the other hole is opposite to the middle of the transparent part of that glass. The back sight vane has one perforation, which is exactly opposite to the middle of the transparent slit in the back horizon glass.

The adjusting lever (fig. 2), which is fixed on the back of the quadrant, serves to adjust the horizon glass, by placing it parallel to the index glass. When this lever is to be made use of, the screw B must be first loosened; and when, by the adjuster A, the horizon glass is sufficiently moved, the screw B must be fastened again; by this means the horizon glass will be kept from changing its position.

To adjust a quadrant.

As the quadrant, from various accidents, is liable to be out of order, it is necessary that the mariner should be able to ascertain the errors, and re-adjust the several parts, before he proceeds to make his observations. For this purpose, he must examine whether the index glass and the horizon glasses be perpendicular to the plane of the instrument, and whether the plane of the fore horizon glass be parallel, and that of the back horizon glass perpendicular to the plane of the index glass, when 0 on the vernier stands against 0 on the limb.

1st. To ascertain whether the index glass be perpendicular to the plane of the quadrant.

Place the index on the middle of the arc, and hold the index glass near the eye. Look into it, in a direction parallel to the plane of the instrument, and see if the reflected arc appear exactly in a line with the arc seen direct, or if the image of any point of the arc near B appear of the same height as the corresponding part of the arc near C seen direct; if so, the index glass is perpendicular to the plane of the quadrant; if not, the error must be rectified by the screws on the base, behind the frame, by loosening the screw M, and tightening the screw N, or by loosening the screw N, and tightening the screw M.

2d. To ascertain whether the fore horizon glass be perpendicular to the plane of the quadrant.

Having adjusted the index glass, hold the instrument in a vertical position. Lock through the fore sight vane, and move the index till the reflected and direct images of the horizon, seen in the horizon glass, coincide. Then incline the instrument till its plane is nearly parallel to the horizon; if the images still coincide, the horizon glass stands perpendicular; otherwise it does not, and must be adjusted by the screws placed before and behind it, loosening one of them, and tightening the other.

This adjustment may be made by the sun, moon, or a star, by holding the quadrant in a vertical position, and observing if the object seen by reflection appears to the right or left of the object seen direct, and moving the screws, as above, till both images coincide.

After having made the horizon and index glasses parallel, according to the directions in the following article, it will be best to re-examine this adjustment.

3d. To make the horizon glass parallel to the index glass, when 0 on the vernier stands

on 0 on the arc.

Having fixed the index, so that 0 on the vernier stands on 0 on the arc, look at any distant object, and see if the image of it coincides with the object itself; if it does, the

adjustment is complete; if not, they must be made to coincide by means of the adjusting lever. The horizon may be used for this purpose in the following manner:Hold the plane of the instrument vertical; look through the lower hole in the vane K, and direct the sight through the transparent part of the glass G to the horizon; then if the horizon line, seen in the silvered and transparent part, coincides, or makes one straight line, the horizon glass is said to be adjusted; but if the horizon lines do not coincide, slacken the screw B (fig. 2) in the middle of the adjusting lever, and turn the horizon glass on its axis until the horizon lines coincide; then fix the lever firmly by tightening the screw B. If this adjustment be again examined, it will perhaps be found imperfect. In this case, therefore, it remains either to repeat the adjustment, or find the error of it (usually called the index error), which may be done thus:-Let the horizon glass remain fixed, and move the index till the image and object coincide; then the difference between 0 on the vernier and 0 on the arc is the index error, which is to be added to the angle or altitude observed, if the 0 on the vernier be to the right hand of 0 on the arc, otherwise to be subtracted. Thus, if the horizon is used, the instrument being held in a vertical position, you must look through the lower hole of the vane K, towards the horizon; then move the index till the reflected and direct images of the horizon coincide; the difference between 0 on the vernier and 0 on the arc will be the index error.

4th. To adjust the back horizon glass, that it may be perpendicular to the plane of the index glass, when 0 on the vernier stands on 0 on the arc.

Set the index as far to the right of 0 on the arc, as twice the dip of the horizon (taken from Table XIII.); hold the quadrant in a vertical position; look towards the horizon through the hole in the back horizon vane L, and the transparent slit of the back horizon glass H; then, if the reflected horizon, which will appear inverted, coincide with that seen direct, the glass is truly adjusted; otherwise the screw, in the centre of the lever on the under side of the quadrant, must be slackened, and the glass turned on its axis till both horizons coincide, when the lever should be fixed by tightening the screw.

5th. To adjust the back horizon glass, that it may be perpendicular to the plane of the· quadrant.

Put the index on 0; hold the quadrant nearly parallel to the horizon; look through the hole on the back sight vane, and if the true and reflected horizons appear in the same straight line, the glass is perpendicular to the plane of the instrument; but if they do not coincide, the sunk screws, before and behind the glass, must be turned till both appear to form one straight line.

To take an altitude of the sun by a fore observation.

If the sun is bright, turn down one or more of the dark glasses; hold the instrument in a vertical position; apply the eye to the upper hole in the fore sight vane, when the image is so bright as to be seen in the transparent part of the fore horizon glass, otherwise to the lower hole; direct the sight to that part of the horizon beneath the sun, and move the index till you bring the image of his lower limb to touch the horizon directly under it; but as this point cannot be exactly ascertained, the observer should move the instrument round to the right and left a little, keeping, as nearly as possible, the sun always in that part of the horizon glass which is at the same distance as the eye from the plane of the quadrant;* by this motion the sun will appear to sweep the horizon, and must be made to touch it at the lowest part of the are; the degrees and minutes pointed out by the index, will be the observed altitude of the sun's lower limb at that instant.

To take an altitude of the moon by a fore observation.

In the night, when the moon is bright, her image may be seen in the transparent part of the fore horizon glass, and the observation may be taken exactly in the same

In common quadrants, if the upper hole be looked through, the sun's image must be made to appear in the middle of the transparent part of the horizon glass; but if the lower hole be looked through, the image must be made to appear on the line joining the silvered and transparent parts of the horizon glass, as these parts of the horizon glass are at the same distances from the plane of the instrument, as

manner as an observation of the sun. If the image is so faint as not to be seen in the transparent part of the horizon glass, you must set the index to 0%; hold the plane of the quadrant in a vertical position; direct the sight to the moon, and, at the same time, look for her reflected image in the silvered part of the horizon glass; move the index forward till the moon's image (which will appear to descend) just touches the horizon; then sweep the quadrant as in observing the sun, and bring her round limb in contact with the horizon, whether it be her upper or lower. The degrees and minutes pointed out by the index, will be the observed altitude of that limb which was brought in contact with the horizon.

To take an altitude of a star by a fore observation.

This is done exactly in the same manner as in observing the moon's altitude, when her image is so faint as not to be seen in the transparent part of the horizon glass.

To take the sun's altitude by a back observation.

Put the dark glasses in the hole O, and turn one or more of them down, according to the brightness of the sun; then, holding the instrument in a vertical position, look through the back sight vane towards that part of the horizon opposite the sun; move the index till the sun's image is seen in the silvered part of the glass; give the quadrant a slow vibratory motion, and the sun will appear to describe an arc with its convex side upward; bring the upper limb, when in the upper part of this are, in contact with that part of the horizon seen through the transparent slit, and the degrees and minutes pointed out by the index will be the altitude of the sun's lower limb. The altitude of the moon, or a star, may be obtained in the same manner, only observing to bring the round edge of the moon to the horizon.

The back observation is but little used, on account of the difficulty of adjusting and observing. Various remedies have been proposed for these defects, but none have vet been generally adopted. The back observation of the altitude of any object, is useful only when there is not an open horizon for the fore observation; but even in that case, the fore observation may often be used, if the distance of the horizon be known, as will be explained hereafter.

To observe the meridian altitude of any celestial object by a fore observation.

When the object rises and sets, it comes to the meridian above the horizon only once in 24 hours, and is then at its greatest altitude; and by observing it, the latitude may be easily determined. The sun comes to the meridian exactly at noon, or 12 o'clock apparent time; the moon and stars at various hours. To observe the meridian altitude, begin a few minutes before the time of passing the meridian; bring the object to sweep the horizon, according to the preceding directions; this operation must be repeated until the object begins to descend below the edge of the sea; the degrees and minutes then shown by the index will be the meridian altitude.

If the object does not set, it comes to the meridian below the pole, and is then at its least altitude; this altitude may be observed as above directed, with this difference, that you must continue sweeping till the object begins to rise above the edge of the sea, instead of descending below it.

The meridian altitude of any object may be taken in a similar manner by a back observation.

Strictly speaking, this method of finding the meridian altitude is not absolutely accurate, except the ship be at rest, and the sun's declination constant. For if the ship is sailing towards the sun, the altitude will be increased; but the altitude will be decreased in sailing from the sun. The correction of altitude arising from this source is generally very small, and it may be neglected in most cases, as will be shown hereafter.

Advice to seamen in the choice of a quadrant.

The joints of the frame must be close, without the least opening or looseness, and the ivory on the arc inlaid and fixed, so as not to rise in any place above the plane of the instrument; all the divisions of the arc and vernier must be exceedingly fine and straight, so that no two divisions of the vernier (except the first and last) coincide, at time, with the divisions of the arc. All the glasses belonging to the quadrant.

the same

should have their surfaces perfectly plane, and their fore and back surfaces exactly parallel; this may be verified, in the horizon glass and index glass, by means of two distant objects, in the following manner :-Move the index till both objects are exactly in contact, at the upper edge of the silvered part of the horizon glass; then move the quadrant in its own plane, so as to make the united images move along the line, separating the silvered from the transparent part of the horizon glass; and if, in this motion, the images continue united, the reflecting surfaces are good planes, otherwise the planes are imperfect. To examine the dark glasses, we must bring the image of a distant object to coincide with the object seen directly; then turn the colored glass so that the plane which was next to the index glass may now be next to the horizon glass, and if the direct and reflected images still coincide, the surfaces of the glass are parallel.

DESCRIPTION AND USE OF A SEXTANT OF REFLECTION.

A SEXTANT is constructed on the same principles, and may be used for measuring altitudes in the same manner, as a quadrant. The arc of a sextant, as its name implies, contains 60°, but, by reason of the double reflection, is divided into 120°. This instrument is particularly intended to measure the distance of the moon from the sun, a planet, or a fixed star; and as that distance is wanted as accurately as possible, to determine the longitude of the place of observation, the instrument is constructed with more care, and is provided with some additional appendages that are not in the quadrant. Plate IX., figure 3, represents a sextaut, the frame being generally made of brass, or other hard metal; the handle at its back is made of wood. When observing, the instrument is to be held with one hand, by the handle, while the other hand moves the index. The arc AA is divided into 120°, each degree into 3 parts of 20 minutes each, and the vernier scale is in general so divided as to show half or a quarter of a minute. In some sextants, the degree is divided into six equal parts, of 10 each, and the vernier shows 10".

In order to observe with accuracy, and make the images come precisely in contact, a tangent screw B is fixed to the index, and by this it can be moved with greater regularity than it can be by hand; but the screw B does not act until the index is fixed by the screw C, at the back of the sextant. Care must be taken not to force the tangent screw, when it arrives at either extremity of its arc. When the index is to be moved any considerable quantity, the screw C must be loosened; and when the index is brought nearly to the division required, the back screw C must be tightened, and then the index moved gradually by the tangent screw.

In many sextants, the lower part of the index glass, or that next the plane of the instrument,silvered as usual, and the back surface of the upper part painted black; a screen, раши Mack, is fixed by its axis to the base of the index glass, and may be placed over the s. vered part when the rays are strong; in this case, the image is to be reflected from outer surface of the upper part, and the error which might possibly arise from the planes of the glass not being parallel, is thereby avoided.

The colored glasses are similar to those applied to a common quadrant, and are usually four in number, placed at D, to screen the eye from the solar rays, and the glare of the moon; they may be used separately or together, as occasion requires. In addition to these, there are three similar glasses, placed behind the horizon glass, to be used in finding the index error by means of the sun, and in observing the sun's cititude, by an artificial horizon on land. The paler glass is sometimes used in observing altitudes at sea, to take off the strong glare of the horizon below the sun, arising from the sun's light, reflected irregularly from the small rippling waves-an appearance which has lately been called kumatage.

A sextant is generally furnished with a tube without glasses, and two telescopes, the one representing the objects erect or in their natural situation, the other inverting them,

*There is not, in general, any apparatus for the back observation fixed to a sextant; but if the altitude of any celestial object be greater than 60°, the supplement of the altitude may be obtained by a back observation, with a sextant, with ease and accuracy; and as this method may be often used with advantage, when a fore observation cannot be obtained, we shall here point out the method of taking the observation, and shall hereafter give the calculations for determining the latitude from a meridian observation, taken in this manner:-The back of the observer being turned to the sun, he must Tove the index till the image of the sun touches the edge of the back horizon, and then move the sextant a little to the right and left (as in a fore observation), and the image will describe an are with the convex side upward; move the index till the lower limb of the image, when in the upper part of the arc, just Conches the horizon, and the observation will be complete; observing that, if the telescope be used, the nage must be brought in the middle between the two parallel wires; but if the telescope be not used, the image of the sun must be seen in the horizon glass, at the same distance from the plane of the strument as the eye of the observer. The altitude thus obtained will be the supplement of the altitude f the sun's upper limb. The corrections to be applied to obtain the true central altitude, will be given mereafter.