The hypothesis of continental drift has become of increasing interest to geophysicists in recent years. The IUGG Upper Mantle Committee has stated that the hypothesis of continental drift envisages horizontal displacements of the continents over thousands of kilometers, and that it is a principal objective of the Upper Mantle Project to prove whether or not continental drift has occurred. The origin of the hypothesis may be traced to the close similarity in outlines of the coasts on the two sides of the Atlantic Ocean. The theory that the eastern and western hemispheres are drifting apart was expounded in particular by A. Wegener.

Much of the evidence for continental drift, such as the matching of coastlines, provides only an average rate over a very long time interval. Palaeomagnetic results give average velocities between 0·5 and 3·0 cm/yr, with higher values for some areas such as India. These also represent average rates over long times. Very recent research on the magnetic character of the ocean floors suggest that the present rate of ocean floor spreading can be determined. The continents of South America and Africa may be separating at a rate of 4·0 cm/yr.

The analysis of 66 years of concurrent latitude observations of the ILS shows that the mean pole has a secular motion which consists of a progressive component of about 0′′.0035/yr (10 cm/yr) along the meridian 65°W and a librational component (oscillation) of 24-year period along the meridian 122°W (or 58°E). Crustal displacements in latitude are not found within the errors of observation, about 1 cm/yr.

Comparable, concurrent observations for time (longitude) have not been made but programs are being organized. From 30 to 50 years will be needed for detection of continental drift with PZT's and astrolabes if relative drifts in longitude of 3 cm/yr are occurring.

Analyzing the notable increases in the residual latitudes of the five ILS stations, local drifts of −′0′.00156/year for Mizusawa and of +0′.00105/year for Ukiah were derived. Subtracting the apparent motion of the mean pole due to the local drifts, the secular motion of the mean pole is 0′′.00220/year in the direction 77°.7 W and this seems to be the real one.

It is necessary to standardise the definition of the direction of the minor axes of spheroids used for geodetic computations. In view of polar motion, the definition “parallel to the axis of rotation” is ambiguous. Few countries have explicitly stated the direction adopted, and it has generally been implicitly defined by the corrections for polar motion applied to astronomical azimuths, latitudes, and longitudes. But these corrections have often been neglected or applied inconsistently, and errors of perhaps 0′′.5 of azimuth result. The choice for a recommended Geodetic Mean Pole, to which spheroidal minor axes can in future be defined to be parallel, seems to lie either with the Cecchini “new system, 1900–05” or with the Mean Pole of Epoch 1962·0.

The relation between the secular variation in latitude and the secular motion of the pole is discussed.

The secular variations in longitude differences among several observatories were determined, but it is hard to consider that these results are literally due to continental drift. The large standard deviations of the data of longitude differences show the existence of external systematic errors in time observations.

The ILS results reduced to a uniform system give a motion of the pole of 0′′.0032/year in the direction 70°W. The latitude of Mizusawa decreases 0′′.003/yr and that of Ukiah increases at the same rate. This confirms the geological theory which indicates a counter-clockwise rotation of the coasts of the Pacific.

The reduction of time service results on a uniform system permits the secular determination of longitudes with high precision. A long interval of time is needed to eliminate quasi-periodic effects. One notes that the variations are similar when stations are in the same terrestrial region and opposed when in different regions.

In developing new methods of computation for the BIH we reached the following conclusions: (1) it is necessary for the BIH to use polar coordinates referred to a fixed origin; (2) the measurement of continental drifts by stations on nearly the same latitude may be carried out with PZT's. When the latitudes are very different, astrolabes can provide useful results.

Three aspects of the experimental study of earth-tides present a direct interest for the study of the rotation of the earth. These are (1) the experimental determination of the Love numbers and the dynamical effects of the liquid core on nutations of the principal axis of inertia; (2) the delay of earth-tides; and (3) the drift of horizontal pendulums.

The coordinates of the pole were derived from both time and latitude observations made at eight observatories from 1962 to 1964. The differences between these results and the ILS results show an annual variation. After removal of this term the standard deviation is about ±0′′.03 for the x- and y-components. Corrections to the adopted longitudes and latitudes are obtained and these agree closely with those obtained by the BIH.

The relations between paleomagnetic observations, polar wandering curves, and changes in the speed of rotation of the Earth are discussed.

It has been proposed to the U.S. National Aeronautics and Space Administration that optical retro-reflector packages be placed on the lunar surface under either the Surveyor or Apollo Programs. Methods for measuring the range to the reflectors with an expected accuracy of 15 cm have been presented. The new technique is briefly discussed, and an analysis of the determination of geocentric longitude is given, indicating a potential uncertainty of 0·25 × 10−3s of time.

Radio tracking of artificial satellites has provided longitude differences between stations on different continents to 10 m. An accuracy of 1 m in the future is expected.

To determine the position of the instantaneous pole one may use, in addition to calculations, a graphical method, which is described.

Explanations of the progressive and librational motions of the pole are attempted.