With the increasing number of surveys based on astronomical coordinates and with loran stations established on oceanic islands far from the mainland and its geodetic control, there has been an increasing interest in deflections of the vertical. Some of those who have a professional interest in deflections of the vertical but who have not analyzed the underlying conceptions seem to have the vague idea that there is something in nature that might properly be called the deflection of the vertical. In the present state of geodesy there is nothing that may properly be called the deflection of the vertical at a given point. All existing numerical values of the deflection depend on the assumed geodetic datum, and the geodetic datum might conceivably be anything whatever, though in practice the choice is confined within rather narrow limits.