The Secular Acceleration of the Moon

if we assume that Atomic Time is a uniform time scale over a few decades, then we may analyze the behavior of t1'e mean longitude of the moon against that time scale. Previously, the moon's mean longitude was used to determine Ephemeris Time; and systematic trends were undetectable, being completely ab- sorbed into time scale. An extensive analysis of occultations of stars by the moon between 1955 and 1969, since Atomic Time has been available, indicates a residual secular acceleration of (- 15"+2") T2 (T in centuries), in addition to the previously assumed -11 `.`22T tidal acceleration determined by Spencer- Jones, and the +7'.' 14T2 dynamical acceleration determined by Brown. Possible causes of this residual acceleration include: (i) additional tidal acceleration; (ii) long-period deficiencies in the lunar theory; (iii) secular variation in length of the Atomic second relative to the Ephemeris second. Analysis of possible sources of systematic error affecting this result seems to imply that the total acceleration not yet accounted for theoretically (including the tidal acceleration) lies between - 18" Th and -34" T2. This does not exclude the possibility that the acceleration has remained constant over the past 30 centuries, as judged by recent results from studies of ancient observations.