The development of self-registering meteorological instruments began very shortly after that of scientific meteorological observation itself. Yet it was not until the 1860's, two centuries after the beginning of scientific observation, that the self-registering instrument became a factor in meteorology.

            This time delay is attributable less to deficiencies in the techniques of instrument-making than to deficiencies in the organisation of meteorology itself. The critical factor was the establishment in the 1860's of well-financed and competently directed meteorological observatories, most of which were created as adjuncts to astronomical observatories.