The gravitational lensing equations for convergence, potential, shear, and flexion are simple in polar coordinates and separate under a multipole expansion once the shear and flexion spinors are rotated into a "tangential" basis. We use this to investigate whether the useful monopole aperture-mass shear formulae generalize to all multipoles and to flexions. We re-derive the result of Schneider and Bartelmann that the shear multipole m at radius R is completely determined by the mass multipole at R, plus specific moments Q ^(m) _in and Q ^(m) _out of the mass multipoles internal and external, respectively, to R. The m ≥ 0 multipoles are independent of Q _out. But in contrast to the monopole, the m < 0 multipoles are independent of Q _in. These internal and external mass moments can be determined by shear (and/or flexion) data on the complementary portion of the plane, which has practical implications for lens modeling. We find that the ease of E/B separation in the monopole aperture moments does not generalize to m ≠ 0: the internal monopole moment is the only nonlocal E/B discriminant available from lensing observations. We have also not found practical local E/B discriminants beyond the monopole, though they could exist. We show also that the use of weak-lensing data to constrain a constant shear term near a strong-lensing system is impractical without strong prior constraints on the neighboring mass distribution.