First, the circle must be placed exactly at right angles to the axis of the telescope, so that it is in the plane of the meridian. Secondly, the error of centering must be found. For instance, if the Greenwich circle were to be read by only one microscope, an error in the pivot or any part of the axis round which the circle turns would vitiate the readings; but we could get rid of that error, due to a fault of the axis, or to a want of centering, by means of two readings, at the extremities of a diameter; but even then we should not get rid of the possible error due to graduation, for even if the divisions on the circle were accurate at first, they would not long remain so, for the metal of which these circles are made is liable, like other metals, to certain changes due to temperature; and if a circle is very large the weight of the circle itself, supposing its form perfect when horizontal, will, when vertical, sag it down and deflect it out of shape, so that at Greenwich the method adopted is to use six reading microscopes. Fig. 120, which shows the Cambridge Transit Circle, indicates the arrangement of the five microscopes in use there, set round the circumference of the circle, much in the same manner as in the case of the Greenwich instrument, where there are holes through the pier in which the microscopes are placed with the eye-ends arranged in a circle at the side of it. [Illustration: FIG. 120.—Cambridge (U.S.) meridian circle. ] When, therefore, the transit is pointed to any particular star, not only is the time noted in order to determine the right ascension of the star, in a careful and elaborate way, but the readings of the circle are made by every one of these microscopes—reading from the next five minutes division of the circle which happens to be visible,—and there is an additional microscope giving the rough reading of the larger divisions of the circle from a certain zero. And what, then, is this zero? There is no doubt about the reading of the zero of right ascension, it is the intersection of the two fundamental planes at the first point of Aries; but what zero shall be used in the case of the vertical circle? [Illustration: FIG. 121.—Diagram illustrating how the pole is found. ]