A very important step for getting absolute proper motions is the extragalactic calibration of the proper motions. We have used three different kind of methods. The first method is based on one bright extragalactic object in the centre of our field. Since we don't expect to measure any real motion of such distant objects, we may assume that the measured proper motion represents only a zero point error of our system. Therefore we have to subtract the apparent proper motions of the extragalactic object from all proper motions for the calibration. This method is denoted with EO in Col. 5 of Table 1 (click here). The internal errors of our proper motions give a lower limit to the accuracy of the calibration of our proper motions. In each coordinate we obtained 1 to 2 mas/a for our objects. The drawback of this method is, that the use of only one object for the calibration does not allow a correction for a possible rotation etc. of the proper motions in the field. Unfortunately none of our plates contains more than one bright extragalactic object, which would improve the link of the proper motions as demonstrated by Odenkirchen & Brosche (1995).
The second method uses stars measured on plates of the Lick astrograph partly from the Lick Northern Proper Motion (NPM) program (Klemola et al. 1987) as described in detail in Geffert et al. (1997). 16 (NGC 4147) to 85 (NGC 7089) stars were used to calibrate our proper motions to an extragalactic reference frame by the determination of a constant offset of our proper motions with respect to the ones obtained from the Lick plates. The rms of the differences were 5 mas/a for each field, which is in good agreement with the quoted errors for the NPM program (Klemola et al. 1987). Four of our fields were calibrated by this method, which is denoted by LICK in Col. 5 of Table 1 (click here). According to (Brosche et al. 1991b) the zero point uncertainty of this link method is of the order of 1.3 mas/a.
For the third method galaxies and stars were measured on Palomar glass copies and ESO Schmidt plates in the same manner as done by Geffert et al. (1993) for the calibration of proper motions in the field of M 15. This method was applied only in the NGC 7089 field. The zero point accuracy of the proper motions is of the order of 1.3 mas/a. In the field of OQ 208 and NGC 7089 we were able to perform two independent extragalactic calibrations of the proper motions. It is very satisfying to note that in both fields the results of both methods agree on the level of 2 mas/a.