The Cassini Division in Saturn's rings contains a series of eight named gaps, three of which contain dense ringlets. Observations of stellar occultations by the Visual and Infrared Mapping Spectrometer onboard the Cassini spacecraft have yielded 40 accurate and precise measurements of the radial position of the edges of all of these gaps and ringlets. These data reveal suggestive patterns in the shapes of many of the gap edges: the outer edges of the five gaps without ringlets are circular to within 1 km, while the inner edges of six of the gaps are eccentric, with apsidal precession rates consistent with those expected for eccentric orbits near each edge. Intriguingly, the pattern speeds of these eccentric inner gap edges, together with that of the eccentric Huygens Ringlet, form a series with a characteristic spacing of 006 day^-1. The two gaps with non-eccentric inner edges lie near first-order inner Lindblad resonances (ILRs) with moons. One such edge is close to the 5:4 ILR with Prometheus, and the radial excursions of this edge do appear to have an m = 5 component aligned with that moon. The other resonantly confined edge is the outer edge of the B ring, which lies near the 2:1 Mimas ILR. Detailed investigation of the B-ring-edge data confirm the presence of an m = 2 perturbation on the B-ring edge, but also show that during the course of the Cassini Mission, this pattern has drifted backward relative to Mimas. Comparisons with earlier occultation measurements going back to Voyager suggest the possibility that the m = 2 pattern is actually librating relative to Mimas with a libration frequency L 006 day^-1 (or possibly 012 day ^-1). In addition to the m = 2 pattern, the B-ring edge also has an m = 1 component that rotates around the planet at a rate close to the expected apsidal precession rate (?? ?? ?? _B ??? 5.??06 day ^-1). Thus, the pattern speeds of the eccentric edges in the Cassini Division can be generated from various combinations of the pattern speeds of structures observed on the edge of the B ring: ??_p = ?? ?? ??_B -jL for j = 1, 2, 3, ???, 7. We therefore suggest that most of the gaps in the Cassini Division are produced by resonances involving perturbations from the massive edge of the B ring. We find that a combination of gravitational perturbations generated by the radial excursions in the B-ring edge and the gravitational perturbations from the Mimas 2:1 ILR yields terms in the equations of motion that should act to constrain the pericenter location of particle orbits in the vicinity of each of the eccentric inner gap edges in the Cassini Division. This alignment of pericenters could be responsible for forming the Cassini-Division Gaps and thus explain why these gaps are located where they are. ?? 2010. The American Astronomical Society. All rights reserved.