Report of the Working Group on Satellites for the period 1981-1984 (Y. Kozai) a) ASTROMETRIC OBSERVATIONS Rohde, Ianna, Stayton and Levinson (31.099.089) published astrometric observations for satellites of the outer planets made in 1978-1979, 1980 and 1981 at the Leander McCormick Observatory, a total of 1032 pairs of spherical-equatorial coordinates and 1316 intersatellite positions being reported. Debehogne and Houziaux (32.099.111) gave positions for the Galilean satellites observed in April 1979 at ESO. Ianna's group is continuing their astrometric observations of brighter satellites of Jupiter, Saturn, Uranus and Neptune at the McCormick Observatory and with the Yale-Columbia refractor at Mt. Stromlo. Bocsa started astrometric observations of Jupiter and its Galilean satellites, photographing the five bodies with the 0.38m-astrograph at Bucharest Observatory; observations made in 1982 have been reported. Harrington and Walker (37.101.021) reported positions of satellites of Uranus and Neptune relative to their respective planets during oppositions of 1979, 1981, 1982 and 1983. Their plates were taken with the 1.55-m reflector in Flagstaff, the planet being occulted with neutral density spots for magnitude compensation. The positions of the satellites of Uranus were sent to the Jet Propulsion Laboratory for use in planning for the 1986 flyby mission. Pascu made photographic observations of satellites of Mars, the Galilean satellites of Jupiter, and Saturnian satellites I through VII with the 0.66-m refractor and with the 1.55-m reflector at Flagstaff. Pascu, Seidelmann, Baum and Schmidt(32.031.546, 34.091.048) observed Jupiter XIV, Saturn XII, XIII and XIV and the satellites of Uranus and Neptune using the Space Telescope Wide- Field Planetary Camera Ground-Based CCD system on the 1.55-m reflector. Veillet and Martins (IAU Circ.3940, 1984) reported their observations of Saturn XII and the Tethys Lagrangian satellites, S XIII and S XIV on 1984 April 12-19 with the Danish 1.5-m reflector at ESO and gave the following data for the orbital separations : S XII-Dione = + 74¡.6 (36 measurements), S XIII-Tethys = + 61¡.3 (11 measurements) and S XIV-Tethys = -65¡.0 (25 measurements), the standard deviations being roughly 0¡.6. The following observations made in the USSR were published : observations of the Galilean satellites in 1976-1978 with the normal astrograph at Pulkovo by Bronnikova and Kiseleva (32.099.151); observations of the Galilean satellites and bright satellites of Saturn in 1976-1979 at the Nikolaev Observatory by Voronenko and Gorel' (32.099.007); observations of the Galilean satellites in 1977-1978 at the Engelhardt Astronomical Observatory by Nefed'ev and Chugunov (37.099.072); observations of the Saturnian satellites in 1977-1978 at the Engelhardt Observatory by Chugunov (37.100.060), and those in 1980 and 1981 by Kitkin and Chugunov (32.100.012, Izv. AEO No 48, 96, 1984); observations of the Galilean satellites in 1975 at the Goloseevo Observatory (Kiev) by Levitskaya (31.099.051); observations of Deimos in 1978 at Goloseevo Observatory by Major and Sereda (Ref. Zh. Astr. 3.51.115, 1984). Several attempts have been made to separate Charon, a possible satellite of Pluto, from the planet. For example, Reitsema, Vilas and Smith (34.101.019) reported that images obtained with a CCD detector attached to the 1.54m-Catalina telescope of the University of Arizona on 1980 February 3 showed an elongation caused by the satellites. Analysis of the data separated the planet and satellites components, yielding a Pluto/Charon brightness ratio of 5.5. Hege et al. (32.101.002) and Hege and Drummond (IAU Circ. 3986, 1984) have applied the technique of speckle interferometry to observations of the Pluto-Charon system and reported corrections to the ephemeris of Harrington and Christy (29.101.011). Speckle interferometric observations have also been reported by Hetterich and Weigelt (34.101.007). Mulholland and Binzel (37.101.020, A.J. 89, 1759, 1984) discussed the expected eclipse phenomena, which had not yet started in 1983. b) MOTIONS OF SATELLITES Mutual phenomena of Galilean and Saturnian satellites have been used to improve their orbital elements. Aksnes et al. (37.099.008) gave astrometric data derived from photoelectric observations of mutual occultations and eclipses of the Galilean satellites in 1973 and 1979/1980 and of four Saturnian satellites in 1980, with accuracy approaching 0".01. Orbital elements based on these data and derived by using radii of satellites measured by the two Voyager spacecraft are in generally good agreement with other recent analyses. Some of the observations used were reported by Arlot et al. (32.099.001). Dourneau (32.100.007) observed two mutual events of Saturnian satellites, S III partially occulting S IV on 1980 April 5, and S V partially eclipsing S III on 1980 April 20, finding that the two events occurred, respectively, about 1 min and 3 min in advance of the predictions by Aksnes and Franklin (1978). It was concluded that a small difference could exist between observed and theoretical longitudes for S III, Tethys, which was involved in both events. Aksnes and Franklin (Ctr. for Ap. Preprint No 1987, 1984) gave predictions for nearly 300 observable mutual eclipses and occultations of the Galilean satellites expected between 1985 May and 1986 April. Predictions for the 1985-1986 events were given also by Arlot (Astr. Ap. 138, 113, 1984). Arlot, Morando and Thuillot (Astr. Ap. 136, 142, 1984) rediscovered an important set of old observations of the Galilean satellites that form a part of the eclipse observations collected by Delambre; some 845 mutual phenomena of Io between 1775 and 1802 are iincluded in their list. Lieske (34.099.028) examined the Delisle manuscript and the Pingre book and recorded and reduced the observations listed there. He now has more than 6800 eclipse observations of the Galilean satellites before 1800 and over 16 000 prior to 1982 and has shown the usefulness of such observations for the adjustment of the theory, particularly for the correction of the mean motions. Tsuchida, Ferraz-Mello and Biancale (31.099.090) discussed the accuracy of photographic observations of the Galilean satellites of Jupiter in the interval 1913-1928 and showed that the residuals are generally very small, 0".06 to 0".08 in the mutual distances. Arlot (31.099.038) calculated new constants for the Sampson-Lieske theory of motion of the Galilean satellites by using a set of 8856 photographic observations; comparison was made with observations of mutual phenomena in 1979. Thuillot and Vu (34.099.112) reported the development of a first approximation of an analytical theory for the Galilean satellites using Sagnier's method, the results being reduced to numerical tables of longitude, radius vector and latitude for easy comparison with Sampson's or Lieske's tables. Rocher (33.099.052) computed ephemerides of Jupiter's satellites, J VI and J VII, for the years 1981-1990 by a numerical integration method and by fitting observational data analogous to that used by Bordovitsyna and Bykova(1978). Boronenko and Schmidt (Tomsk University) continued to work on a theory of the motion of distant satellites using the Lie transformation technique; they plan to apply it to Jupiter VI, VII, X and Saturn IX. Bec-Borsenberger and Rocher (32.100.083) collected the topocentric observations of Phoebe (S IX) from 1904 to 1981 and obtained initial integration constants adjusted with these observations. They then made a numerical integration up to 1990, taking into account perturbations by the Sun and Titan, and computed an ephemeris for the ten years 1981-1990. Bykova and Shikhalev (37.100.019) analyzed observations of 1898-1981 and derived new orbital elements of Phoebe which represent observations with a mean error of 1".5. Zhang and Liu (32.100.104) surveyed the theory of motion of Hyperion (S VII). Taylor (Astr. Ap., in press, 1984) used astrometric observations of Hyperion from 1967 to 1982 to derive values of parameters of Woltjer's theory for a best fit to the modern data. He found, however, that there are significant inadequacies in the theory. Accordingly, Sinclair and Taylor have tried to fit a numerical integration of the motions of Titan, Hyperion and Iapetus to the modern data, obtaining a much better fit to the Hyperion data and a good determination of various physical parameters of the system. Yoder et al. (33.100.047) presented a simple analytical theory describing the 1:1 orbital resonance and applied it to the Saturnian co-orbiting pair, 1980 S1 (S X) and 1980 S3 (S XI). These small satellites can approach within 15 000 km, but they are prevented from passing each other by their mutual gravitational interactions. Long-term stability was also discussed, and a tie was established between the 1966 and 1980 observations of the two satellites. Chugunov (32.100.016, 33.100.060, 33.100.061, Ref. Zh. Astr. 11.51.65, 1983; Ref. Zh. Astr. 12.51.73, 1983; Ref. Zh. Astr. 4.51.266, 1983; Ref. Zh. Astr. 4.51.120, 1984; Ref. Zh. Astr. 6.51.179, 1984) determined new orbital elements and masses of Saturnian satellites S I - S VI as well as the mass of Saturn and the values of J2 and J4 by using 28 500 observations. Thuillot (34.099.032) proposed a new method to derive compact sets of ephemerides of various phenomena of the Galilean satellites by calculating a Chebyshev representation; 15 coefficients give a representation of each type of phenomenon for one satellite for one year with a precision better than 51 seconds. The Bureau des Longitudes , Paris, publishes each year in the Connaissance des Temps the positions of the planets and the four Galilean satellites of Jupiter developed into Chebyshev polynomials, and in its Supplements the configurations and phenomena of the Galilean satellites, the positions of the first eight satellites of Saturn, and the positions of J VI, J VII, J VIII, J IX and S IX developed also into Chebyshev polynomials. Veillet (33.101.004) gave 112 positions of Miranda relative to Uranus in 1980-1981 and confirmed an inclination which agrees with the value found by Whitaker and Greenberg from observations in 1948-1972. However, the eccentricity derived is very small. A more extended study, using Van Biesbroeck's data from 1948-1949 as well as all published positions of Miranda (300 points), confirmed all the orbital elements obtained except for those linked to the eccentricity, for which the value found was 0.0021 ± 0.0005. The masses of Ariel and Umbriel were derived by using the values of J2 and J4 derived from the apsidal precession rates of the rings and the nodal precession rate of Miranda. Veillet (32.101.005) used a series of observations of Nereid made with the Danish-ESO 1.5-m reflector in 1981 April, together with two unanalyzed plates taken at the McDonald Observatory in 1977 and 1978 to determine orbital elements of Nereid and the mass of Neptune. c) THEORETICAL STUDIES OF SATELLITE MOTIONS Most theoretical studies of satellite motions have been devoted to the question of dynamical evolution of the orbits, considering such topics as capture mechanisms and the effect of resonances. Tanikawa (33.042.024) proved the impossibility of the capture of retrograde satellites in the frame of the circular planar restricted problem of three bodies. Huang and Innanen (34.042.037) explored numerically the stability and capture region for retrograde jovicentric satellites in the frame of the restricted three-body problem and derived the value of the Jacobian constant for the greatest probability of temporary capture. Szeto (34.097.008) examined the role of tidal dissipation within the Martian satellite system and assessed theories of origin through calculations of collision probabilities between Phobos and Deimos in the distant past. An accretion model is preferred over capture, although no such model consistent with the likely carbonaceous chondritic composition of the Martian satellites has yet been established. Henrard (33.099.020) reewamined Yoder's scenario (25.099.047) for the capture into resonance of the first three Galiean satellites by introducing a more refined dynamical model for resonance and tidal effects. Gailitis (32.099.099) showed that the limited energy of the radial motion has preserved Io from a runaway type of melting as the energy-momentum balance relates tidal heating to orbital expansion as in Yoder's secular equation. Sinclair (34.100.023) critically reviewed the hypothesis that the origin of the resonances among the Saturnian satellites is due to orbital evolution caused by tidal dissipation within the planet. He concluded that the hypothesis provides a plausible explanation for the origin of the Mimas-Tethys resonance, but it is unsatisfactory for Enceladus-Dione, since their resonance now has little effect on the relative evolution rate of these satellites. Poirier, Bolah and Chambon (34.100.014) investigated tidal dissipation in a viscoelastic homogeneous sphere having the orbital and physical characteristics of the icy inner satellites of Saturn and found that tidal dissipation with current orbital eccentricity cannot account alone for the surface activity observed on Enceladus if it is composed of water ice. Lissauer, Peale and Cuzzi (37.100.051) argued that the angular momentum transfer between Enceladus and Janus could have sufficiently enhanced the eccentricity of Enceladus'orbit for tidal heating to have melted the interior of Enceladus if Janus were ever locked into a stable 2:1 orbital commensurability with Enceladus. The rapid time scale for dynamical evolution of the ring and inner satellites as presently situated remains a major problem. Farinella et al. (33.100.089) proposed that the peculiar orbital motion of Hyperion, characterized by a strong orbit-orbit resonance with Titan, was responsible for the ineffectiveness of the process of repeated reaccretion from narrow rings of collisional fragments proposed for the other satellites. The irregular shape of Hyperion thus stands in contrast to the regular figures of the other small satellites of Saturn. Pauwels (34.100.005) studied the Rhea-Titan secular resonance as a special case of secular orbit-orbit resonance. Taking into account both masses, it was shown that in the Rhea-Titan case, the resonance is dominated by the effect of the large difference between the proper precession rates on the lines of apsides of the two satellites. Sinclair (Astr. Ap. 136, 161, 1984) examined the effects of orbital resonances on satellites in tadpole or horseshoe orbits relative to Mimas, Enceladus, Tethys or Dione and concluded that a tadpole companion of Enceladus would be the most highly perturbed. Although the perturbations would not cause instability, they might prevent the initial formation of a satellite in such an orbit. Tadpole companions of Mimas would have large forced eccentricities, which would probably prevent the simultaneous existence of horseshoe orbits. He also examined the effects of tidal forces, which were found to cause a small displacement of the L4 and L5 equilibrium points. d) DYNAMICS OF RINGS Study of the dynamical processes that cause the complex structural features of rings, such as those manifested by the narrow elliptic rings of Uranus and the F ring of Saturn, as well as the spokes in the Saturnian rings, has been an extremely active field during the triennium. Reviews have been published by Goldreich and Tremaine (32.091.025) and by Cuzzi (37.091.052). IAU Colloquium No 75, held in Toulouse in September 1982, was devoted entirely to the subject of planetary rings. The proceedings of the colloquium, containing contributed papers and also the discussion, have been edited by Brahic and published in 1984 under the title Anneaux des Planetes/ Planetary Rings by Cepadues-Editions on behalf of CNES, the French space agency. A second book, based on a set of review papers selected from among those presented at Colloquium N¡ 75, has been edited by Greenberg and Brahic and published in 1984 under the title Planetary Rings in the Space Science Series of the University of Arizona Press. The material in this second book has been carefully reviewed and refined, and reflects the results of investigations published in 1983. Since full lists of literature citations are given in both books, mention is made here of only a few of the most recent studies. Borderies, Goldreich and Tremaine (34.042.039) extended their investigations of the evolution of eccentric rings under the influence of differential precession due to the oblateness of the planet, self-gravity, viscous forces due to interparticle collisions, and eccentricity excited by shepherding satellites. They concluded that uniform precession can be enforced by self- gravity, the resulting configuration being both secularly and dynamically stable, that due to viscous forces the line of the apse at the inner ring edge is not exactly aligned with the line of apsides at the outer edge, the apse shift being detectable in the alpha and beta rings of Uranus, and that the mean eccentricity is determined by a balance between viscous damping and excitation by shepherding satellites. Lissauer (37.100.002) has studied ballistic mass transport in Saturn's rings, developing an analytic model that includes the effects of angular momentum advection. He showed that the net material movement due to angular momentum advection is comparable to that caused by direct ballistic transport. Seidelmann, Harrington and Szebehely (37.100.052) studied the dynamics of Saturn's E ring, which extends from three Saturn radii, the orbit of Mimas, to at least eight Saturn radii, just inside the orbit of Rhea. Although the brightness profile has a peak at the orbit of Enceladus, there are apparently not corresponding peaks at the orbital distances of Tethys or Dione. It is known that there are satellites at the Lagrangian points of Tethys and Dione within the E ring. Thus the E ring is not constrained by satellites, but rather co-exists at the same orbital distances as satellites, which have other satellites at their Lagrangian points. A discussion of preliminary numerical and analytical investigations of the motions of the ring particles is given.