The subject I have chosen for my talk today will come of age on the last day of this General Assembly. The first detection of the remarkable objects that we call pulsars, was on 28th November 1967 when Jocelyn Bell (now Bell-Burnell) who was working with Prof. Hewish at Cambridge discovered a new class of radio sources that put out pulses of radio radiation about once a second but with clock-like regularity. Two other major discoveries in the same decade using the radio spectrum were, of course, Quasars and the Cosmic microwave background. Those two took us to extremes in time and distance, and the amount of energy radiated, whereas the discovery of pulsars, situated near at hand by comparison, led us to extremes in the physical state of matter.

Thirty years ago, observational cosmology consisted of the search for two numbers: Ho, the rate of expansion of the universe at the position of the Galaxy; and qo, the deceleration parameter. Twenty years ago, the discovery of the relic radiation from the Big Bang produced another number, 3oK. But it is the past decade which has seen the enormous development in both observational and theoretical cosmology. The universe is known to be immeasurably richer and more varied than we had thought. There is growing acceptance of a universe in which most of the matter is not luminous. Nature has played a trick on astronomers, for we thought we were studying the universe. We now know that we were studying only the small fraction of it that is luminous. I suspect that this talk this evening is the first IAU Discourse devoted to something that astronomers cannot see at any wavelength: Dark Matter in the Universe.

I feel very honoured at having been asked by the IAU Executive to present Prof. Sagdeev’s invited discourse on Venus in his absence. As has already been mentioned Prof. Sagdeev was summoned to the talks between the USSR and the USA in Geneva as an expert in Space Research. He went there with several very important proposals for future cooperation in scientific exploration of outer space and we all hope that the results of his mission will be beneficial for the development of space astronomy and therefore for the whole astronomical community represented by the International Astronomical Union. I am sure that all present in this room realize that it is not an easy task to present someone else’s lecture especially at such short notice and in a foreign language. Besides, it is my opinion that a talk on Venus – the brightest celestial body in the sky except the moon and the Sun – Venus which always has been the symbol of beauty, love and femininity, that such a talk should surely be given by a man.

Galactic astronomy requires mainly accurate proper motions of stars, referred to a system which represents an inertial system as closely as possible. We discuss the required accuracy, give an outlook on forthcoming reference frames of use for galactic astronomy, and discuss observationally determined upper limits on the rotation of the extra-galactic reference frame.

Because of relatively large orbital inclinations of some minor planets and many comets, coupled with occasional close approaches to the Earth, it may be necessary to measure positions of these objects anywhere in the sky. Brightnesses range widely, some comets becoming bright enough to be readily visible with the naked eye, while small or remote objects challenge the limit of detectability at the largest telescopes. The brightness of a given object observed over an interval of time also will vary, sometimes over many magnitudes, as geocentric and heliocentric distances change.

It is suggested that proper motions will continue to play a key role in fixing the extragalactic distance scale and that uncertainties in the present reference frame are of significance for distance scale problems.

Observations of Earth orientation parameters (polar motion and Universal Time) have reached a level of accuracy which challenges the definition of the reference systems which these observations seek to relate. Observational accuracies of better than in polar motion and in Universal Time are now routinely available. Urgent requirements therefore exist now for the IAU to take a fresh look at the concepts and definitions of the reference systems as well as reference frames. Of particular concern are the definitons of the equinox, the terrestrial reference pole, the zero of longitude and sidereal time.

When solving some abstract problems in mechanics related to the dynamics of bodies and systems, the notion of an inertial frame of reference is introduced in an apparently clear and natural way by simply drawing its coordinate axes and then paying no further attention to the system of reference which is then taken for granted. If we turn, however, towards investigations of the real stellar and planetary world, or as Sir James Jeans put it, “… to the Universe around us,” we immediately face the question of how to practically construct a useful and obvious model of the inertial frame of reference sufficiently close to reality.

A survey is given on the tasks to be performed in the process of improving a fundamental system. Completed tasks are the derivation of corrections to the values for general precession and the determination of the FK4 equinox and equator. The selection of new fundamental stars and the improvement of the systematic and individual accuracy of the FK4 is in progress. A description of the observational material and of new methods of analysis is given. Preliminary results are reported.

It was recommended in a resolution of the International Astronomical Union that as of Jan. 1, 1984, a new astronomical reference frame would be adopted (IAU 1976). This reference frame has been designated as the FK5 reference frame of J2000.0. The problem of the interim realization of this reference frame pending completion of work on the compilation of the FK5 catalog at the Astronomisches Rechen-institut has been addressed by a number of authors, see, e.g., Kaplan(1981), Standish(1983), Aoki et al.(1983), and Lederle and Schwan(1984). Their discussions are primarily concerned with the steps which must be taken to transform FK4 catalog positions and proper motions referred to the equator, equinox and epoch of B1950.0 to J2000.0. The primary emphasis of the following discussion is on the transformation of catalogs other than the FK4 to the J2000.0 equinox. There are a few remarks concerning the FK4 catalog.

In astronomy we try to determine a non-rotating frame from analyses of the observed motions of three mechanical systems – the solar system, the galaxy and the extragalactic nebulae. The rotation of the extragalactic frame is of the order 10-10 arcsec per century, so, for all practical purposes, this frame may be regarded as having no rotation. The other two frames are model-dependent and, as such, cannot be regarded ab initio as constituting non-rotation frames of reference. These reference frames are linked by various techniques, as shown in the diagram below.

The MERIT programme of international collaboration to monitor earth-rotation and to intercompare the techniques of observation and analysis has fostered the development of the use of space techniques. Earth-rotation parameters are now determined regularly with a precision that is better than 1 milliarcsecond () and the relative positions of the observing stations are determined to better than 1 decimetre (0.1 m). It is therefore necessary that the terrestrial and celestial reference frames be defined more precisely. The MERIT and COTES Working Groups have proposed that new conventional terrestrial and celestial reference systems be established and that the maintenance of these systems be the responsibility of a new International Earth Rotation Service. The new reference frames are to be based on the adoption of positions and motions of designated stations and extragalactic radio sources. Appropriate models and parameters will be associated with these frames to form reference systems so that observations can be used to determine the rotation of the terrestrial frame with respect to the celestial frame.

A discussion of four methods that can be used to link HIPPARCOS to an absolute reference system shows that the best method is to refer it to extragalactic objects. This would free the resulting frame from any parameter describing Earth’s motion.

The Hubble Space Telescope (HST) will have the capability of making relative astrometric measurements with an accuracy of four to ten times better than similar ground based measurements. The instruments of choice will be the Fine Guidance Sensors (FGS’s). Details of the instrumentation have been given elsewhere (c.f. Jefferys, 1980). Each FGS consists of two interferometer detectors which look at a small aperture located within a larger field of view (FOV). The raw data consist of encoder readings, from “star selectors”, and error signals derived from the interferometer transfer functions. The star selector positions determine the position of the small aperture within the pickles. The precision is always less than 0.001 arcsec, with an expected accuracy of 0.002 to 0.003 arcsec per observation. Under optimum conditions, the best expected accuracy is 0.0016 arcsec per observation. An observation will consist of the relative positions of several objects in one pickle. Thus the highest accuracy observations will be relative observations within a relatively small field of view.

The Lick Northern Proper Motion (NPM) and the Yale-San Juan Southern Proper Motion (SPM) programs have been described on several occasions (Wright 1950; Deutsch and Klemola 1974; Vasilevskis 1973; and Wesselink 1974). The two programs represent an attempt to measure the coordinates and proper motions, with respect to the extragalactic reference frames, for large numbers of stars representing most of the astrophysically-recognized classes. The photographic plate material forming the basis of the NPM program derives from the first (1947-1954) and second (1970-present) epoch phases for 1246 fields with the Lick 51 cm Carnegie double-astrograph for centers at -20° and northward (Shane and Wirtanen 1967). A southern supplement of 144 additional fields takes the program to -30°. The SPM consists of 632 fields with centers at -20° and southward. The first epoch plates were taken between 1965 and 1974 and a partial extension of 72 fields to more northerly declinations is essentially complete. One short and one long exposure permit the measurement of positions and approximate photometry for selected stars and reference galaxies over the blue magnitude range from about 8 to 17-18.

In recent years, a revolution in astronomical position measurements has been taking place with the advent of modern space techniques. These new techniques, which supplement the traditional astrometric measurements, include laser ranging to the moon and artificial satellites, very-long-baseline interferometry (VLBI) of galactic and extra-galactic radio sources and spacecraft, radio tracking of satellites, and radar-ranging and spacecraft tracking during planetary encounters. Impressive accuracies have been achieved and further improvements are forthcoming. Each technique can be expected to establish its own reference frame which is derived from observations of a particular class of objects. The celestial and terrestrial coordinate systems are related through adopted constants and definitions. Contemporary astronomy has led to the development of three principal celestial coordinate systems: the optical frame (FK4/FK5) based on positions of galactic stars; the planetary/lunar ephemeris frame based on the major celestial bodies of the solar system; and the radio frame constructed from observations of extragalactic radio sources (quasars). Each frame is rotated with respect to others; furthermore, the optical frame offset is time variable. It is important that all frames be interconnected and unified. The optical frame is being connected to the radio frame by VLBI observations of radio emitting stars. The radio frame is being tied to the ephemeris frame in several ways – one is via differential VLBI measurements between quasars and planet-orbiting spacecraft.

The catalogues of positions of radio-sources obtained by VLBI, their link with precise and dense star catalogues (HIPPARCOS) and the atomic time open an era where “natural” references in space and time will be quasi-ideal. Are our traditional concepts and usages adapted to this new situation?

Astrometry can be defined as the measurement of space-time coordinates of photon events. For example, in principle, in classical optical astrometry, we measure the components of velocity, and hence the direction, of an incoming photon with respect to an instrumental coordinate system, and the clock time, at the instant of detection. The observer’s coordinate system at any instant can be identified with a local inertial frame. In the case of interferometric observations, the measurements are of clock times of arrival of a wavefront at two detectors whose spatial coordinates are specified with respect to instantaneous inertial frames.

The Very Large Array (VLA) has made possible the measurement of the precise positions of the radio emission associated with stars. This allows the direct comparison of the optical reference frame (FK4) with the radio reference frame which is defined by the quasi-absolute positions of extragalactic radio sources. This comparison is limited by the small number of bright stars that display detectable radio emission and the lack of knowledge of the precise coincidence of the radio emission with the optical photocenter of the star. Since the VLA is the most sensitive astrometrically capable radio telescope, positions of the largest number of stars north of declination -20 degrees can be measured. The accuracy of the positions on the extragalactic reference frame should approach a milliarcsecond.