VOL. XLVI.] PHILOSOPHICAL TRANSACTIONS. 6p' 



halves ; which may be easily estimated into smaller parts if necessary. The di- 

 visions are determined by means of a cat-gut line stretched perpendicularly with 

 the box as near the brass edge as may be, that the parallax arising from a different 

 position of the observer may be as little as possible. Under the card are 2 small 

 weights, sliding on 1 wires, placetl at right angles to each other ; which, being 

 movetl nearer to or farther from the centre, counterbalance the dipping of the 

 card in different latitudes, or restore its equilibrium, where it happens by any 

 other means to be got too much out of level. 



There is also added an index at the top of the inner box, which may be put on 

 and taken off at pleasure, and serves for all altitudes of the object. It consists 

 of a bar, equal in length to the diameter of the inner box , each end being fur- 

 nished with a perpendicular stile, with a slit parallel to the sides. One of the 

 slits is narrow, to which the eye is applied, and the other is wider, with a small 

 catgut stretched up the middle of it, and from thence continued horizontally 

 from the top of one stile to the top of the other : there is also a line drawn 

 along the upper surface of the bar. These four, viz. the narrow slit, the hori- 

 zontal catgut thread, the perpendicular one, and the line on the bar, are in the 

 same plane, which disposes itself perpendicular to the horizon, when the inner 

 box is at rest, and hangs free. This index does not move round, but is always 

 placed on so as to answer the same side of the box. 



When the sun's azimuth is desired, and his rays are strong enough to cast a 

 shadow, turn about the wooden box, till the shadow of the horizontal thread ; 

 or, if the sun be too low, till that of the perpendicular thread in one stile, or 

 the light through the slit in the other, fall on the line on the index bar, or vi- 

 brate to an equal distance on each side of it, gently touching the box, if it vibrate 

 too far : obsei-ve at the same time the degree marked on the brass edge by the 

 catgut line. In counting the degree for the azimuth, or any other angle reck- 

 oned from the meridian, make use of the outer circle of figures on the brass 

 edge ; and the situation of the index bar, with regard to the card and needle, 

 will always direct on what quarter of the compass the object is placed. 



But if the sun do not shine out sufficiently strong, place the eye behind the 

 narrow slit in one of the stiles, and turn the wooden box about, till some part of 

 the horizontal or perpendicular thread appear to intersect the centre of the sun, 

 or vibrate to an equal distance on each side of it ; using smoked glass next the 

 eye, if the sun's light be too strong. In this method another observer will be 

 generally necessary to note the degree cut by the nonius, at the same time the 

 first gives notice that the thread appears to split the object. From what has 

 been said, the other observations will be easily performed ; only in the case of 

 the sun's amplitude, take care to number the degree by the help of the inner 



