A method of successive approximations in the theory of isotope shifts, consisting in a series expansion with respect to charge powers of the effective nuclear radius of interaction, is applied.
The second order contributions prove to be sufficiently weak. More important seem to be the effects yielded by the shape of the proton distribution. In addition, the possible effect of triaxiality is considered.
A reformulation of hydrodynamic corrections shows that surface energy gives no contribution and the symmetry energy brings an almost negligible contribution, as a result of the statistical variational model for the nucleus.
The interpretation of the existing empirical data in terms of our theory — made in the last section of this paper — leads to a quite satisfactory agreement.