The successful improvements of meridian instrumentation during the past twenty years have mainly come through electronic devices and computers being applied for facilitating the data handling and for increasing the accuracy. Photographic recording of the circle and of the star has played an important role here, but has for both tasks gradually lost its attractiveness as direct photometric-electronic methods have become available. At the same time new types of the telescope system have been introduced without convincing results as yet, but this field still holds promises for the near future.
The problems of the refraction, the meridian building, the foundation of the instrument and the site selection have been treated in recent years and these efforts will be doubly repaid when the telescope and micrometers become more nearly perfect. Altogether, it need not be long before a few partly automatic instruments produce observations with a mean error of 0″.15 and a systematic error of 0″.03 at a rate of 300 observations per observing night. In addition the limiting magnitude can be mv
= 11 or 12, thus 2 mag. fainter than for visual observations. These goals are conservative – most of them have already sometimes been surpassed – and they should be compared to a present day good visual meridian circle giving a mean error 0″.30, a systematic error 0″.10 and 120 observations per night.
The relative roles of meridian and photographic astrometry must be defined anew in the light of the great improvements of both methods.