Acknowledgement. We thank Science for their permission to use an excerpt from:
Smith, B. A., et al. 1981. Encounter with Saturn: Voyager 1 Imaging Science Results Science 212 (4491), 163-191. (Excerpt from pp. 182-190.)
The rings. Voyager 1's observations of Saturn's rings reveal a much more complex and varied structure than expected from Earth-based photos or Pioneer 11 images (4). Here we discuss the overall properties of the rings seen in Voyager images: the major components of the system, their large-scale structure, and gross comparisons that can be made among them.
As seen in Fig. 30, the classical ring elements (A, B, C, D, Cassini division, and Encke division) retain their significance in that they differ in the qualitative nature of the observed structure. Closest to Saturn is the extremely faint D ring, seen in Fig. 31. Material in this ring has been detected at a radial distance of about 66,500 km from the planet center, but it could extend down to Saturn's atmosphere; its outer edge is the inner edge of the C ring. The D ring consists of numerous narrow features that vary in width from several hundred kilometers to the resolution limit of 35 km. These features were observed in both forward and back scattering, indicating a negligible component of small (micrometer sized) particles. The optical depth is extremely small and it is very likely, despite reports to the contrary (63), that the D ring has never been observed from the ground.
The C ring consists of broad bands of optically thin regions, giving it a transparent appearance in transmitted light (Fig. 32). Within these transparent regions are many optically thick ringlets, at least one of which is an eccentric feature (see Fig. 33).
The B ring is the largest component of the ring system, extending over 25,000 km, and is in general the brightest and most optically thick. At small phase angles the B ring exhibits modest structure; at large phase angles it resolves into hundreds of bright ringlets and dark gaps, with widths from 100 km down to the limit of resolution. Over most of the B ring, little or no large-scale (several hundred kilometer) systematic spacing of ringlets is evident. When seen in diffuse transmission, many of the narrow features in the ring appear relatively transparent. No evidence for such narrow, randomly distributed, optically thin regions exists outside the B ring. The boundary between the B and C rings is sharp (~ 10 km) and is bordered outward by a 600- to 800-km-wide region of extremely large optical depth, with no evidence of narrow gaps.
The next classical feature, at a radius of approximately 118e3 km, is the Cassini division, which until recently was believed to be a single gap. Voyager 1 high-resolution images (Fig. 34 ) reveal a wealth of detail: five broad, sharp edged, evenly spaced bands, each of which exhibits structure down to the limit of resolution. The arrangement of large and small-scale features in the Cassini division, as well as in the C ring, shows a regularity that at present cannot be attributed to resonances with known satellites. The optical density of the bands is similar to that of features in the C ring. This similarity is paralleled in the photometric properties of these two ring elements, which are notably different from those of the A or B ring (see Fig. 30).
The A ring and the Encke division within it are the last major features of the classical ring system. Although in comparison with the other components the A ring is devoid of spectacular detail, high resolution images show many narrow and well-ordered features that may be associated with satellite resonances. The reported azimuthal asymmetry in A ring brightness has been confirmed in low resolution Voyager images. The Encke division is not entirely free of material but contains at least two very narrow, clumpy ringlets.
Beyond the outer edge of the A ring there are several more features. The F ring, discovered by Pioneer 11 (4), is found to be far more complicated than was imagined. This ring lies between the orbits of 1980S27 and 1980S26 at a radius of about 140,000 km. It will be discussed later.
Still farther out is the G ring at a radius of about 170,000 km. It can be seen in Fig. 31 as a thin line across the frame. Image smear has broadened the ring's appearance; however, the ring seems to be very faint and very narrow. Charged particle experiments on Pioneer 11 (60) indicated the presence of material at this radius.
Figure 31 also shows the E ring in forward-scattered light. This ring is extremely optically thin and extends from about 210,000 to 300,000 km in radius. The E ring's visibility only in forward scattered light supports observations of Terrile and Tokunaga (42) which indicate that it has a large component of small particles .
Photometry of the rings. Only a limited number of the Voyager ring observations have been photometrically calibrated, and thus only preliminary numbers for ring optical depth, particle albedo, and phase function can be presented at this time. Before discussing these results, we give a qualitative overview of the greatly varying appearance of the main rings with sun-ring-spacecraft geometry.
Before encounter, the lit face of the rings was observed at low phase angles (~ 10 degrees to 15 degrees). The roughly equal brightness of the A and B rings and the low brightness of the rings relative to the planet are due to the low solar elevation angle (3.8 degrees at the time of encounter). After Voyager 1 passed through the ring plane to the unlit side, a quite different appearance of the rings was observed (Fig. 35). At this geometry the nearly totally opaque B ring and a variety of finer scale features appear relatively darker than the C ring and Cassini division. Regions of intermediate optical depth, such as the A ring, are intermediate in brightness. Truly vacant regions, such as the inner and outer gaps within the Cassini division and the Encke division, also appear dark due to the nearly total absence of scattering material.
During observations of the unlit face, viewing geometry allowed the rings to be seen against the bright limb of the planet, permitting determination of the optical depth of various regions (Fig. 32). After Voyager recrossed the ring plane and emerged to the lit face, the scattering geometry had changed and the rings were viewed at higher phase angles (130 degrees to 160 degrees). Regions characterised by strongly forward-scattering particles now appeared brighter than when viewed in backscatter before encounter (Fig. 36).
These variations of ring appearance with illumination and viewing geometry under the three general viewing conditions are summarized and illustrated in Fig. 37.
Images of the rings taken through various filters have been calibrated and reduced to I/F units (17), following the calibration procedure described by Soderblom et al. (64). Figure 38a illustrates radial photometric profiles from images taken through a green filter and an ultraviolet filter with a resolution of about 280 km/lp. The intensity scales are proportional to I/F, but have been scaled so that the brightest parts of the B ring are at the same level for both filters. A marked difference in the two profiles is evident; the C ring and the material in the Cassini division are considerably brighter than the A and B rings in the ultraviolet than in the green filter image.
Figure 38b shows relative spectral reflectance at five spectral bandpasses (scaled to unity at green) for portions of the A, B, and C rings and the Cassini division. The A and B rings show essentially the same spectral properties, while the C ring and Cassini division show enhanced reflectance in the ultraviolet and the violet. Also shown in Fig. 38b are ground-based data for the B ring (65). The agreement indicates that the camera characteristics have not changed dramatically since Jupiter encounter. Trial calculations indicate that the large difference in reflectance at shorter wavelengths for the C ring and Cassini division is not due to multiple scattering in regions of different optical depths. It is probably due to intrinsic differences in the optical properties of the ring particles, with the C ring and Cassini division particles having a more neutral (less red) color.
The azimuthal asymmetry of the A ring observed from the ground (66) is also seen in Voyager 1 preencounter images of the ring system. Preliminary measurements on images taken 25 days before encounter at a resolution of 600 km/lp indicate that the modulation of ring brightness has approximately the same amplitude in all colors (blue, green, orange, and violet) and is comparable to the ground-based value, a peak-to-peak variation of ~ 15 percent. The phase angle for the Voyager measurements was 12 degrees, larger than is possible from Earth, with a viewing (ring-opening) angle of 11 degrees. The phase of the variation with azimuth seen from the spacecraft exactly matches that seen from Earth. Minima in brightness occur in postconjunction quadrants at ~ 20 degrees from the ansae and maxima occur in preconjunction quadrants ~ 40 degrees from the ansae.
We made preliminary radiative transfer model calculations of the variation of ring brightness with viewing geometry and compared them with calibrated Voyager data. The radial regions used (Table 4) are not particularly heterogeneous and average values were used. Optical depths were obtained from calibrated data in Fig. 32 by differencing brightness across the limb of the planet (67).
The brightness of the limb was obtained from unobscured images at comparable latitudes and phase angles. No meaningful optical depth can be obtained for the B ring due to the dominance of light transmitted through a small fraction of narrow gaps, as inferred from Pioneer 11 photometry (67). However, its average optical depth is a more relevant quantity for diffuse brightness calculations and is sufficiently large (> 1) to have a negligible effect on brightness calculations for the lit face.
It was realized before encounter (68) that the rings, especially the A and B rings, were more strongly backscattering than previously suspected. A phase function for Lambert spheres was assumed for initial computations, and a doubling code was used which treated the rings as having a bimodal distribution of optical depths, inferred from Pioneer 11 data (67). The small inhomogeneity thus introduced does not significantly affect the model brightness values shown. Typical results are shown in Fig. 39. The low phase angle data were used to obtain particle scattering albedos, which are given in Table 4, and refer to the clear filter of the wide-angle camera, with center wavelength ~ 5070 Angstroms. The value given for the A and B rings is in excellent agreement with that obtained from ground-based observations and analyses (69). The particles in the C ring and Cassini division are significantly darker than the A and B ring particles. The scattering phase function of the A and B ring particles is not satisfactorily matched by the Lambert sphere assumption. This may be seen by comparing the brightness of these regions observed at high phase angles (Fig. 37) with that observed at low phase angles. The high phase angle data fall at significantly lower albedo values, implying that the A and inner B ring particles are significantly more backscattering than Lambert spheres, a result consistent with the strong opposition brightening exhibited by many planetary bodies and particles with porous surfaces. The behavior of the outer B ring is not the same as that of the inner B ring; it is more forward scattering than Lambert spheres. We have not yet attempted more detailed modeling of the particle phase functions. Interestingly enough, the C ring and Cassini division particles agree much better with the Lambert sphere assumption over the whole range of phase and viewing angles. Obviously, much refinement and expansion of these results must occur before detailed conclusions can be drawn about the nature of the particles in various regions.
The C ring. The structure of the C ring is regularly ordered and cannot be readily associated with gravitational resonances. Its inner edge, for instance, is not in correspondence with any predicted resonance location. A prominent gap several hundred kilometers wide at about 77,300 km does agree fairly well in location with an apsidal resonance with Titan (70, 71). Intermediate regions of the C ring show a low-contrast, undulating pattern. In the outer regions, an even more prominent gap ~ 270 km wide is flanked in a regular and symmetrical manner by several relatively featureless bands that are more opaque than their surroundings and have sharp edges. This gap contains a narrow ringlet that is eccentric, as shown graphically in Fig. 33.
The outer gap is at a radius of about 87,100 km and the eccentric ringlet is 35 km wide and 100 km from the outer edge of the gap at its periapse. The ringlet is optically thick when viewed against Saturn's disk. At the apoapse the ring is 90 km wide and 50 km from the outer edge of the gap. The ring has one periapse and one apoapse, indicating that its eccentricity is not induced by resonance with an outer satellite. The line of apsides was not observed to move over a period of several days. This is consistent with a long precession period (~ 24 days) for the eccentric ring, probably due to the quadrupole moment of Saturn. It seems likely that this ringlet must continually have its eccentricity forced by some mechanism, probably by a small satellite or satellites in its vicinity. Such forcing would seem to argue for an external resonance, but this feature is located near only a 2:1 resonance with the tiny satellite 1980S28. This feature will be studied by Voyager 2.
In highest resolution (10 km/lp) images of the unlit face of the rings, the inner C ring exhibits many marginally resolved dark features, which are not as common in the outer C ring. The optical thickness of these features is not yet known. The C ring has also been observed at high phase angle (150 degrees), and does not exhibit any significant brightening as would be expected from substantial quantities of small particles.
The B ring. The extremely sharp boundary between the B and C rings does not show any gap, nor does it correspond in location to any significant resonance. However, the properties of the regions on either side of the boundary differ structurally as well as photometrically.
In general, the B ring has the most extensive and uniform small-scale structure, consisting of many sequences of alternating high and low optical depth, with characteristic radial scales of several hundred to tens of kilometers. In transmission, narrow radial gaps of low optical depth are seen to be spaced irregularly, but with spacing comparable to their own widths of tens of kilometers. In forward scattering, bright bands several hundred kilometers wide are defined much more clearly than in backscatter (Fig. 37). Locally regular structure is much more evident in forward scattering than in backscatter. In fact, these bands are so extensive and numerous as to cause the entire outer B ring on the average to brighten considerably in forward scattering. The inner B ring does not exhibit such forward-scattering behavior.
On a scale of 1e3 to 1e4 km, the B ring is divided into four radial zones of roughly equal width that differ qualitatively in photometric properties. The inner zone contains the most numerous and transparent gaps. Near the innermost edge the ring tends to be more opaque and to resemble the extremely opaque outer third quarter. These similar regions of low transparency have fewer narrow gaps. In general, the most opaque regions have the greatest brightening in forward-scattered light. They also exhibit the bulk of the "spoke" activity, discussed below in greater detail. Here we point out that the spoke features show the same increased brightness and definition as do the regions of the rings with which they are primarily associated, and are therefore characterized by a substantial number of small particles relative to surrounding areas.
Cassini division. The outer edge of the B ring is the inner edge of the Cassini division, and is quite opaque and abrupt to the limit of resolution. The boundary appears to correspond to the location of the 2:1 Mimas resonance, which is a resonance of major strength and has long been believed to be associated with the Cassini division. Figure 34 is a three frame mosaic of the Cassini division, representing a small fraction of the Voyager 1 radial coverage of the entire ring system at ~ 10 km/lp. In general, these observations allow us to understand discrepancies in boundaries assigned to the Cassini division from ground-based, Pioneer 11, and preencounter Voyager 1 observations. A bright band ~ 1400 km wide lies outward of the outermost prominent dark region, which is an essentially vacant gap ~ 170 km wide. This bright region, when viewed from above, is comparable in brightness to the A ring and appears to blend with it. When seen from the unlit face, however, it appears comparable to the C ring in opacity (~ 0.1). Therefore, if matter density is the basic criterion for distinguishing regions, it should be regarded as part of the Cassini division, which was essentially the conclusion reached by the Pioneer 11 photopolarimeter team (4). However, the lit-face brightness argues for particle properties similar to those of the A ring, so that if local particle properties are considered the basic criterion, the region should be regarded as part of the A ring.
A very regular sequence of alternating bright and dark features extends practically throughout this interesting (bright band) region, with interval decreasing outward, typically hundreds to tens of kilometers. This pattern may be evidence for density waves within the ring material, with dark areas of concentration and bright areas of rarefaction. Between the two prominent gaps lie four or five bands of similar appearance, each with fine structure at the limit of our resolution. This set of bands was used to produce the photometric results for the Cassini division discussed earlier. Within the inner of the two gaps is a narrow ringlet similar in width, opacity, and eccentricity to the ringlet in the outer gap of the C ring described earlier. We expect that resonant forcing maintains the eccentricity of this ringlet. Its proximity to the Mimas 2:1 resonance makes it of special interest for observations by Voyager 2. Finally, images obtained at large phase angles show that the Cassini division in general does not have a substantial component of small, forward-scattering particles. In this respect as well, it is similar to the C ring.
The A ring. The inner edge of the A ring is as abrupt as the B/C boundary and as opaque as any region of the B ring. Most of the A ring, however, is of lower optical depth and varies smoothly in brightness. The Encke division, at a radial distance of 133,200 km, is vacant except for two very narrow, faint ringlets that are irregularly clumped. These ringlets are brighter in forward scattering and probably contain a significant fraction of small particles.
Although the Mimas 5:3 resonance is nearby, it corresponds better with one of five other features of great interest in the outer A ring. These features are generally dark in backscatter (Fig. 40), bright in forward scatter (Fig. 36), and optically thick relative to surrounding areas. At high resolution (Fig. 41) they exhibit alternating bright and dark patterns with a spacing of tens of kilometers. It should be noted that there are far more theoretical resonances with the external satellites than there are ring features in this region. Perhaps the superposition of two or more resonances is required for their formation. Four of these features are close to four positions where calculated resonances nearly coincide. The remaining feature shows an unusual and intriguing effect in forward scattering; whereas the other features are bright compared to their surroundings, this feature reverses contrast at the ansa. This suggests vertical relief at this radius, which tentatively corresponds to the 5:3 resonance with Mimas' orbital inclination involving the orbital inclination of the ring particles.
Outward of the Encke division, there is a pattern of unresolved A ring features that are dark in backscatter, bright in forward scatter, and decrease in interval with increasing radius. These features may represent a sequence of classical resonances converging on one of the close-in satellites. The entire region outside the Encke division is ~ 25 percent brighter than the region inside; this is echoed by a region at the outer edge of the A ring, several hundred kilometers wide, that is set off from the A ring by an apparently vacant gap tens of kilometers wide and is ~ 50 percent brighter in forward scatter than the region immediately inside the gap (see Fig. 37). The outer edge of the A ring is sharp at the limit of our resolution.
The F ring. Pioneer 11 discovered the narrow F ring about 3600 km outside the edge of the A ring (4). As Voyager 1 approached Saturn two satellites were discovered in close association with the F ring (Fig. 42). Outside the ring, 1980S26 moves in an eccentric orbit that takes it between 2000 and 500 km from the ring. Just 500 km inside the ring is 1980S27, which is comparable in size to 1980S26 (about 200 km in diameter). The F ring was also found to be eccentric, with at least a 400-km variation in its 140,000-km radius. It is not known whether the precession of satellite orbits and F ring eccentricities will change the close-approach distances of the F ring and 1980S26 and 1980S27.
In high-resolution images, bright clumps are seen in the F ring. Their orbital motion, determined by tracking from frame to frame, coincides with the F ring's Keplerian period of about 15 hours. Still higher resolution [Fig. 43 (left)] shows that the F ring has three components, each about 20 km wide. The inner component is the faintest; it is smooth and does not appear to interact with the other two. Perhaps the most startling view acquired by Voyager 1 is shown in Fig. 43 (right). Over half the image, the outer two components of the F ring appear to be intertwined or braided. There are five crossover points with a separation of 700 +/- 100 km (mean and standard error of four measurements). The radial extent of the braids is about 30 km. Throughout the rest of the image no braids are seen, although a series of kinks or knots in the rings is apparent. Nine knots have been measured with a mean separation of 630 +/- 150 km. It is not clear whether the coincidence between the separation of crossover points and the separation of knots is significant. If it is, then ~ 700 km seems to be the characteristic scale length for warps in the F rings. The complex nature of the F ring braids and clumps is most likely connected with the gravitational actions of the shepherding" satellites, 1980S26 and 1980S27. The limited viewing geometries and times did not permit three dimensional characterization of the features or their time evolution. These parameters, to be measured by Voyager 2, and details of the satellites and ring eccentricities are needed to describe F ring gravitational dynamics. Why the third (faintest) component appears smooth and without knots or kinks still remains to be explained.
At high phase angles, the F ring and particularly the ring clumps become brighter, indicating a significant component of small particles. We estimated the (cross section-weighted) average radius r-bar of the particles by comparing the brightness at phase angles of 140 degrees and 155 degrees. Assuming that the F ring is optically thin, it suffices to consider only single scattering. Also, Mie theory may be used, since the single-scattering phase function is insensitive to particle shape at these small angles. We find that r-bar is several tenths of a micrometer. This is much smaller than the particle radii of centimeters to meters that characterize much of the main rings (72).
The small value of r-bar for the F ring particles has several implications. First, as in the case of Jupiter's ring (73), the lifetime of such small particles may be much less than the age of the ring system due to micrometeoroidal collisions and/or sputtering by radiation belt particles. Ejecta from cratering of larger bodies in the F ring or from collisions between these hypothetical bodies (perturbed gravitationally by 1980S26 and 1980S27) may represent a plausible source for the small F ring particles. Second, the small value of r-bar may allow electromagnetic as well as gravitational forces to influence the motion of the F ring particles. Electromagnetic forces would result from charging of the particles by the ambient plasma and the photoelectric effect, with Saturn's corotating magnetic field exerting a Lorentz force on them. Such forces could play a role in determining the strange morphology of the F ring.
Spokes. Among the most interesting variable ring features are the narrow radial markings or spokes in the B ring. These features were first observed as dark wedge-shaped projections rotating in the outer half of the ring, predominantly on the morning ansa (the part of the ring just emerged from Saturn's shadow). At high resolution, spokes were easily observed on both ansae and across most of the B ring. Spokes have not been seen to cross the Cassini division and seldom extend farther in than about 105,000 km from the center of Saturn. This point appears to be a sharp cutoff for most of the spokes and is a bright region (in backscattered light) in the B ring. Characteristic lengths and widths of these features are 10,000 and 2,000 km, respectively.
What were discovered as dark markings during inbound passage (backscattered light) became bright markings (Fig. 44) during outbound passage (forward scattered light). Viewed against the planet's disk, the spokes show no optical density variations, and no spokes are visible from the unilluminated side of the rings. In the highest resolution images of the spokes taken in forward-scattered light, the B ring resolves into radially aligned areas of enhanced brightness along ringlets. This is especially true in the outer optically thin region of the ring. Only in the optically thick regions do the gaps appear to be filled in by spoke material. These observations indicate that small particles constitute a large fraction of the B ring and are more visible, perhaps because they are elevated above the ring plane, within the spokes.
The inferred behavior of small particles in the spokes and the presence of a rotating magnetic field in the ring environment suggest a possible connection between electromagnetic forces and the dynamics of spoke rotation. Spoke motion was measured by comparing images of the rings taken over 30-minute intervals. Figure 45 shows data from two sets of images, one inbound and one out bound, on which spokes are seen on the morning ansa. Rotation rate is plotted against radial distance from Saturn's center and curves representing pure Keplerian motion and the rotation of Saturn's magnetic field are shown for comparison .
These data seem to lie along the Keplerian line, most convincingly at radial distances greater than ~ 10.7e4 km. Although large morphological changes in the spokes are not apparent over an interval of 30 minutes, the diffuseness of these features makes exact determination of their location difficult, particularly at smaller radii, where they are more diffuse. Still, it appears that along most of the length of a spoke, motion is independent of the magnetic field.
The difference between Keplerian and magnetic orbital motion grows as one moves away from a radius of 112.5e3 km (Fig. 45). Hence it is interesting to note that the spokes commonly appear wedge shaped, wider toward Saturn, and that the location at which they become the narrowest coincides with the point where Keplerian orbital motion and magnetic angular velocity are identical. Also, examination of many images indicates that the vast majority of spokes are near-radial or tilting away from radial in such a way that Keplerian motion will continue to tilt them further. These observations suggest that the magnetic field is responsible for creation of the spokes and that Keplerian motion is responsible for their particle dynamics. The characteristic wedge shape may reflect the difference between Keplerian and magnetic orbital motion over the time it takes one spoke feature to be created.
The modulating influence of Saturn's magnetic field could also account for the random occurrence of spokes on the rings. The Voyager 1 radio astronomy experiment (9) detected electrostatic discharge signatures that are believed to originate in the rings and have a modulation period similar to the magnetic field rotation rate. Spokes may be the visible manifestation of these phenomena. Correlation of radial alignment of the spokes with a particular magnetic longitude has not yet been investigated.
Finally, the spokes do not seem to have any common orbital longitude of origin in the ring system. Spoke alignment with the radial direction does not seem to be correlated with any particular orbital location. When pure Keplerian motion was assumed and spoke features were "mapped" backward in time to check for alignment with the morning or evening edge of Saturn's shadow on the ring plane, the features were found not to be parallel to the shadow.
Discussion. Out of the great complexity of Saturn's rings, we can only begin to draw general conclusions. Qualitative differences exist between the classical main ring elements. The C ring and Cassini division are similar in color and albedo and unlike the A and B rings in these properties. Although there are possible mechanisms for varying the surface properties of ring particles as a function of optical thickness (thickness variations could result in shading variations), there does not seem to be an unambiguous pattern in the data. The rings also differ in abundance of small, forward-scattering particles, with the outer B ring, the F ring, the outermost A ring, and the discrete features surrounding the Encke division having the highest relative abundance. The Cassini division and the C ring show the least evidence of such small particles, and they show less backscatter relative to Lambert spheres than do the A and B ring particles. The low abundance of small particles in the C ring and Cassini regions implies that the bluer color of these regions is an intrinsic compositional effect and not due to small-particle Rayleigh scattering behavior. The A and C rings lack the disordered structure evident in the many narrow transparent zones in the B ring; they exhibit a relatively smooth background punctuated by discrete features, sometimes located at classical resonances (A ring) and sometimes not (A and C rings). The boundaries between the rings seem to represent real differences in particle properties. The greater abundance of small particles in certain regions implies local differences in rates of production, as small particles are short-lived. The F ring and outer A ring are likely to be locations of relatively more intense collisions due to the perturbing activity of 1980S26, 1980S27, and 1980S28. Thus it is not surprising to find many small particles there. The B ring exhibits both a preponderance of small particles and a preponderance of narrow gaps. Both phenomena are consistent with a relatively large number of large (~ 1 km) moonlets in the B ring. Such moonlets will clear gaps ~10 km wide around themselves by superposition of torques (71) in a time approximately equal to the diffusion collapse time of such gaps. The gravitationally deposited energy produced by such large particles will then produce large random-collision velocities (74) and consequently greater production rates of small particles. The spoke patterns, representing increased visibility of small particles, are restricted to regions that are already rich in small particles. In the A ring, the fact that many small particles are seen in association with the regularly spaced features near the Encke division is consistent with the idea that the features are density waves, forced by gravitational resonances with 1980S26 and 1980S27 or some other objects, that become nonlinear as they propagate from their source, producing vigorous collisions and numerous small particles. The many unresolved features outside the Encke division may be of similar origin.
Several significant questions remain. If the presence of small, forward-scattering particles is indicative of vigorous local dynamics, why are such particles absent Tom the Cassini division, the location of some of the strongest predicted dynamical activity? If the simultaneous presence of many narrow gaps and many small particles, widely dispersed, is indicative of the presence of many embedded and invisible kilometer-size moonlets, how does one understand the presence and predominance of such objects in the B ring? How are the distinct and abrupt boundaries between the main rings themselves, and the significant qualitative differences in particle properties between the classical regions, maintained? At present, these boundaries are not satisfactorily correlated with any possible classical gravitational resonances, except for the boundary between the outer B ring and the inner Cassini division. While the sharp outer edge of the A ring is understood, the sharp inner edge of the C ring is not.
