724 concomitant with the magnetic field signature, has been described by Sittler et a/ 16 • As additional correlative obser- vations become available, we hope to offer a more detailed model. Of the models in Table 1, cases 2 and 3, derived using weights inversely proportional to the local field magnitude, minimize the effect of the localized current system encountered near periap- sis. This unmodelled 10-nT perturbation is less likely to affect significantly the internal field coefficients in the weighted least- squares fits compared with the unweighted cases, as larger errors are tolerated in higher fields. Case 3 in Table 1 is regarded as best representing the global characteristics of the saturnian field, as deduced from Voyager 1 observations alone. To estimate model parameter uncertainties we need to assess the impact on the internal field coefficients of the unmodelled current system. Lacking a detailed model, we can estimate the parameter uncertainties by comparing models relatively unaffected by the currents (weighted fits) with those that are more strongly dependent on the residuals near closest approach (unweighted fits). Such a comparison suggests a dipole moment of 0.21±0.005 tilted 1.0±0.3° towards an SLS longi- tude of 340° ± 20°. The northward displacement of the dipole by 0.04 Rs deduced from the Pioneer 11 observations is not Received 13 May; accepted 10 June 1981. I. Acuii a, M. H. et a/. Science 207,444-446 (1980). 2. Smith, E. J. eta/. Science 207,407-41 0 (1980). 3. Acuiia, M. H. et at. J. geoph ys. R es. 85, 5675- 5678 (1 980). 4. Smith, E. J. eta/. J. geoph ys. Res. 85, 5655-5674 (1980). 5. Kaiser, M. L. et a/. Science 209, 1238- 1240 (1 980). 6. Ness. N. F. eta /. Science 212,211-217 (1981 ). 7. Acuna, M. H. & Ness, N. F. in Magnetospheric Particles and Fields (ed . McCormack, B. M.) 3 11 -323 (Reidel, Dordrecht. 1976). 8. Davis , J., Jr & Smith, E. J. in Magnetospheric Particles and Fields (ed. McCormack, 8. M.) 30 1-310 (Reidel, Dordrecht, 1976). Nature Vol. 292 20 August 1981 substantiated by the Voyager 1 observations, although we must take into account that the near-equatorial trajectory of Pioneer 11 was poorly suited for detecting hemispherical asymmetries. Voyager 2 will fly by Saturn in August 1981; its close encounter trajectory will provide a latitude coverage of- ±30° within 8 Rs and facilitate a more accurate estimate of a possible dipole polar offset. Certain aspects of Saturn's main magnetic field relevant to the planet's interior have been discussed by Stevenson 17 • In parti- cular, the unexpectedly small dipole moment seems to be consistent with the gravitational settling of helium, which leads to a much smaller electrically conducting and convecting region than would be expected of a homogeneous distribution of hydrogen and helium. The IR studies by Voyager 1 have demonstrated such an atmospheric depletion of helium, obtain- ing an 11% helium mass fraction at Saturn as compared with the 19% observed at Jupiter and expected on a cosmochemical basis 18 • The near alignment of Saturn's magnetic and rotational axes is likewise consistent with the differential rotation of the conducting core in Stevenson's model of Saturn's interior 17 • We thank our colleagues on the Voyager team, especially J. Scudder and E. Sittler of the plasma experiment for helpful discussions. We also thank R. Thompson and F. Ottens for the analysis of these data. 9. Connerney. J. E. P. J. geophys. Res. (in the press). 10. Chapman, S. J. &. Bartels, J. Geomagnerism (Oxford Universi ty Press, 1940). 11 . Desch, M.D. &. Kaiser, M. L. Geophys. Res. Letr. 8, 253-2 56 (1981 ). 12. Connerney, J. E. P. era/. Nature 292,724-726 (1981 ). 13. Bohannon, K. W. eta/. Space Sci. Rev. Zl, 235-257 (1977). 14. Acuiia, M. H. era/. J. geophys. Res. (in the press). 15 . Dess1er, A . J . & Vasyliunas, V. M. Geophys. R es. Lett. 6, 37-40 (1979). 16. Sittler, J. C., Scudder, J. C. & Bridge, H. S. Nature 292, 711- 714 (1981 ). 17 . Stevenson, D. J. Science 208,746-747 (1980). 18. Hanel, R. eta/. Science 112, 192-200 (1981 ). Saturn's ring current and inner magnetosphere J. E. P. Connerney, M. H. Acuna & N. F. Ness NASA/ Goddard Space Flight Center , Laboratory for Extraterrestrial Physics, Greenbelt, Maryland 20771, USA The Voyager 1 magnetic field observations at Saturn reveal an equatorial system of (eastward) azimuthal currents, very similar in certain respects to that responsible for the jovian magnetodisk. IN November 1980, Voyager 1 became the second spacecraft to obtain in situ observations of the magnetosphere of Saturn: little more than a year earlier, the Pioneer 11 spacecraft returned the first evidence of Saturn's magnetosphere. Analyses of the Pioneer 11 magnetic field observations 1 ' 2 revealed a dipolar planetary magnetic field of moment 0. 21 G- rather less than that expected on the basis of scaling laws 3 · 4 • [The units G- R are dimensionally equivalent to standard units of magnetic dipole moment, the numerical value having been divided by the cube of the planetary radius.] More surprising was the near alignment of Saturn's rotational and magnetic dipole axes, a unique characteristic among the known planetary dynamos and therefore of central interest in dynamo theory 5 • A preliminary estimate of 0. 7° angular separation of Saturn's rotational and magnetic dipole axes has recently been derived from the Voyager 1 magnetic field observations 6 • Saturn's lack of an appreciable dipole tilt results in an obser- vational limitation insofar as studies of magnetospheric phenomena in situ by spacecraft are involved. Figure 1 shows the trajectories of each spacecraft to encounter Saturn (Pioneer 11 (P11 ) in September 1979 ; Voyager 1 (V1) in November 1980; and Voyager 2 (V2) in August 1981) in a cylindrical (p, z) planet centred coordinate system, which in this case is practically identical to the magnetic equatorial coordinate system. Each spacecraft remains essentially at constant magnetic latitude for a large percentage of the encounter, with only one (V2) or two 0028-0836/81/340724-03$01 .00 (V1, P11) crossings of the magnetic equator. Particularly evident is the rather singular nature of Pioneer 11 's trajectory, which traced a nearly identical path outbound from periapsis as inbound. In contrast to the periodic variations in the magnetic e.----------------------------------. 6 4 -4 -6 VOYAGER 2 TRAJECTORY (APPROXIMATE) Fla.l Trajectories of Pioneer 11, Voyager 1 and Voyager 2 at Saturn in a cylindrical planetocentric equatorial coordinate system. Positions of the major satellites and rings are indicated; satellite diameters are exaggerated by a factor of 10. Stippled region is the region of (model) distributed currents in Saturn's magnetosphere. © 19X 1 Macmillan Journ al s l .td