Figure 2. This figure shows the amount of secondary piston motion required to correct for defocus owing to sag in the secondary spider rods. During the acquisition of these measurements the telescope collimation process was running, therefore, the defocus shown here is the residual sag which is not being corrected for. The solid line shows the fit of to these data and the ledgend shows the functional form of the fit used. The coefficients m1 through m3 should be added to the SecPistCoefs which were in use at the time of these measurements. The telescope temperature was quite constant during the time these measurements were taken. If this were not the case, then some of these variations could be attributable to temperature changes in the truss.