Ring flow kinematic features, such as, e.g., the shape of the narrow r
ings of Uranus or the streamlines of density waves of Saturn's rings,
can be analyzed by fitting ring occultation data with m-lobe shapes wi
th constant elements and constant precession rates. The orbital parame
ters thus obtained (e.g., the semimajor axis, the eccentricity) are us
ually referred to as geometric elements. This approach raises some que
stions of principle: how are these geometric elements defined? Why is
it justified to apply elliptic formulae (in data analyses) and Gauss p
erturbation equations (in theoretical analyses) to these geometric ele
ments? In our previous papers on these issues (N. Borderies and P. Y.
Longaretti, 1987, Icarus 72, 593-603, P. Y. Longaretti and N. Borderie
s, 1991, Icarus 94, 165-170), we have resolved these questions by show
ing that the geometric elements are in fact appropriately chosen epicy
clic elements, and that the epicyclic formulae and perturbation equati
ons are formally nearly identical to the more familiar elliptic theory
. This paper addresses various issues that were not considered in the
previous papers. First, the epicyclic solution is extended to nonequat
orial motions. Second, the lowest order amplitude corrections to the f
requencies of the motion are computed. These corrections allow us to d
erive the correct contributions to order e2J2 and I2J2 to the orbital
precession rates. Also, an expression of a third nonclassical quasi-in
tegral of the motion in a flattened axisymmetric potential is provided
from the work of G. Contopoulos (1960, Zeit. fur Astrophys. 49, 273-2
91). This third integral is a generalization of the energy in radial o
scillations. Finally, the perturbation equations of the epicyclic elem
ents are derived. These results are important for the determination of
weak dynamical effects in planetary ring systems. (C) 1994 Academic P
ress, Inc.