The second adjustment is to make the pivots horizontal. Place a striding level on the pivots and bring the bubble to zero by the set screws of the level, or note the position of it; then reverse the level east for west, and then if the bubble remains at the same place the axis of motion is horizontal, but, if not, raise or lower the movable Y sufficiently to bring the bubble half way to its original position, and complete the motion of the bubble, if necessary, by the level screw until there is no alteration in the position of the bubble on reversing the level. [Illustration: FIG. 115.—Diagram explaining third adjustment, H, R, plane of the horizon; H, Z, A, P, B, R, meridian; A and B places of circumpolar star at transit above and below pole P. ] The third adjustment is to place the pivots east and west. Note by the clock the time of transit of a circumpolar star, when above the pole, over the central wire, and then half a day later when below it, and again when above it; if the times from upper to lower transit, and from lower to upper are equal, then the line of collimation swings so as to bisect the circle of the star round the pole, and therefore it passes through the pole, and further it describes a meridian which passes through the zenith by reason of the second adjustment. This is therefore the meridian of the place, and therefore the pivots are east and west. If the periods between the transits are not equal, the movable pivot must be shifted horizontally, until on repeating the process the periods are equal. In practice these adjustments can never be made quite perfect, and there are always small errors outstanding, which when known are allowed for, and they are estimated by a long series of observations made in different manners and positions. The error of the first adjustment is called the collimation error, that of the second the level error, and that of the third the deviation error. When the errors of an instrument are known the observations can be easily corrected to what they would have been had the instrument been in perfect adjustment. * * * * *