The following, Fig. 159, will show how we proceed if we merely wish to measure a distance, the value of the divisions of the micrometer screw having been previously determined by allowing an equatorial star to transit. It represents the position of the central and the movable wire when the shadow thrown by the central hill of the the lunar crater Copernicus is being measured to determine the height of the hill above the floor of the crater. It has been necessary to let the fixed wire lie along the shadow; this has been done by turning the micrometer; but there is no occasion to read the vernier. [Illustration: FIG. 159.—How the Length of a Shadow thrown by a Lunar Hill is measured. ] Except on the finest of nights the stars shake in the field of view or appear woolly, and even on good nights the readings made by a practised eye often differ, _inter se_, more than would be thought possible. In measuring distances we have supposed for simplicity that we find the distance that one wire has to be moved from coincidence with the fixed wire from one point to another, and theoretically speaking the pointer should point to O on the screw head when the wires are over each other, and then when the wires are on the points, the reading of the screw head divided by the number of divisions corresponding to 1˝ will give the distance of the points in seconds of arc. But in practice it is unnecessary to adjust the head to O when the wires coincide, and the unequal expansion of the metals of the instrument, due to changes of temperature, would soon disarrange it. It is also somewhat difficult to say when the wires exactly coincide, and an error in this will affect the distance between the points. It is therefore found best to only roughly adjust the screw head to O, and then open out the wires until they are on the points and take a reading, say twenty-two; the screw is then turned, in the opposite direction and the movable wire passed over to the other side of the fixed one, and another reading taken, say eighty-two; now the screw has to be moved in the direction which decreases the readings on its head from one hundred downwards, as the distance of the wires increases, so that we must subtract the reading eighty-two from a hundred to give the number of divisions from the O through which the screw is turned, and the reading in this direction we will call the indirect reading, in contradistinction to the direct reading taken at first. So far we have got a reading of twenty-two direct and eighteen indirect, which means that we have moved the screw from twenty-two on one side of O to eighteen on the other side, or through forty divisions, and in doing so the movable wire has been moved from the distance of the two points on one side of the fixed wire to the same distance on the other, or through double the distance required. Therefore forty divisions is the measure of twice the distance, and the