The star having been found, the date and decimal of the year are entered at the top, and a position taken by bringing the thick wires parallel to the stars. A distance—say direct—is then taken, and the degrees of position 170°·1, and divisions of the micrometer screw seventeen, read off with the assistance of a lamp and entered in their proper columns. The micrometer is then disarranged and a new measure of position and an indirect distance taken, and so on. At the end of the readings, or at any convenient time, the zero for position is found by turning the micrometer until the wires are approximately horizontal, and then allowing a star to traverse the field by its own motion, or rather that of the earth, and bringing the thick wires parallel to its direction of motion; this may be more conveniently done by means of the slow-motion handle of the telescope in R. A., which gives one the power of apparently making the star traverse backwards and forwards in the field. The position of the wires is altered until the star runs along one of them. The position is then read off and entered as the zero. In describing the adjustments of the position circle we made the vernier read 0° when the star runs along the wire, for that is practically the only datum line attainable; since, however, the angles are reckoned from the north, it is convenient to set the circle to read 90° when the star runs along the wire, so that it reads 0° when the wires are north and south. Now as positions are measured from north 0° in a direction contrary to that of the hands of a watch, and an astronomical telescope inverts, we repeat the bottom of the field is 0°, the right 90°, and so on; now the reading just taken for zero is the reading when the wires are E. and W., so that we must deduct 90° from this reading, giving 19°·8 as the reading of the circle when the wires were north and south, or in the position of the real zero of the field. Of course theoretically the micrometer ought to read 0° when the wires are north and south, but in screwing on the instrument from night to night it never comes exactly to the same place, so that it is found easier to make the requisite correction for index error rather than alter the eye end of the telescope to adjustment every night. The readings of position must therefore be corrected by the number of degrees noted when the wires are at the real zero, which in the case in point is 19°·8, which may be called the index error.