“…Although an accurate numerical solution for the beam envelope radii is easily obtained for specified beam and quadrupole lattice parameters, approximate analytical solutions continue to be useful for design studies, scaling, cost optimization, and physical understanding. Various analytical methods have been applied to solve these equations during the last twenty five years [3,4,5,6,7], with the degree of error decreasing from about 10% for the early smooth limit approximation to less than 1 % using the small parameter expansions employed by Anderson [6] and Lee [7]. In the present work the error is reduced to less than 0.1% for typical system parameters, but one may question the vdue of this new work since several approximations have been made in deriving the K-V equations, which may produce errors much larger than 0.1%, These approximations include the neglect of third order geometric abeaatians, non-linear components of quadrupole fringe fields, higher order magnetic multipoles, and deviations fiom the assumed flat space charge profile.…”