The results of several long series of latitude observations have been used for a separate determination of the coefficients of some nutational terms in obliquity and longitude.

The derived value of the constant of nutation is essentially smaller—and those of the coefficients of the semiannual and semimonthly terms larger—than the respective theoretical values, determined on the supposition that the earth as a whole is an elastic body. The theoretical value of the ratio of the axes of the nutational ellipse evidently does not need any correction.

On the basis of these data some conclusions of a qualitative character are made on the interaction between the core and the shell of the earth.

A mechanical explanation of the observed irregular rotation of the surface of the earth is proposed. From this probabilistic model the correlations between the apparent fluctuations of the motion of the moon in successive years are derived and a modified least-square method is developed for finding the secular accelerations of sun and moon.

Geophysical considerations favor seeking the cause of the random changes in the earth's rate of rotation in the turbulent motion in the core of the earth. Professor van der Waerden's analysis applies to this interpretation. The observational evidence indicates that the character of the fluctuation curve may be intermediate between that expected on the basis of a theory with frictional couple and one without. If confirmed, this would indicate that two different causes contribute to the changes in the earth's rate of rotation.

In order to lessen the influence of the errors in the right ascensions of stars of fundamental systems on the results of time determinations, it is proposed that the results of time services observing with transit instruments be used for the determination of an independent system of right ascensions of stars. The participation in this work by a maximum number of observatories of different countries is desirable.

Comparison of photographic zenith tube (P.Z.T.) observations with time derived from quartz-crystal clocks during 1951 to 1955 and with cesium standards of frequency during 1955 to 1958 indicates that the seasonal variation is nearly the same from year to year. Lunar-tidal inequalities of semi-monthly and monthly periods with amplitudes of about 0S.001 each were found. A preliminary value of the Love number, k, is derived. Observations made since 1952 with the dual-rate moon position camera are used to derive ΔT = ET – UT. Comparison of the P.Z.T. observations and atomic standards at the National Physical Laboratory and the Naval Research Laboratory shows details of the irregular variation from 1955 to 1958.

In this article is described the measurement of time from a more general point of view than has been taken before. It is shown that mean solar time and ephemeris time are two particular applications of the general principle of measuring time by observing angular motions, a principle that can be applied to any angular motion whatever, provided that an adequate theory of the motion is available.

Atomic frequency standards have been checked by laboratory investigations and by the intercomparison of different standards.

Existing ammonia Masers enable a frequency to be defined in terms of a spectral line with an accuracy of ±1 part in 108 and give a reproducibility under stated conditions of ±2 parts in 1010. The cesium atomic beam standard enables a frequency to be defined in terms of a spectral line with a precision of ±1 part in 1010. Comparison between the U.K. and U.S. by the radio transmissions of standard frequencies and time signals gave an average agreement of a few parts in 1010 but revealed some systematic discrepancies. Direct comparisons between three U.S. standards and the U.K. standard gave agreement to 2.2, 3.2, and 1.5 parts in 1010.