“…In this case the model without F is generally not integrable, yet (if considered in the complex , namely without restricting the dependent variables z n -nor, for that matter, the coupling constants g 2 nm -to be real) it still does feature an open, hence fully dimensional, region in its phase space where all solutions are completely periodic with the same period T , see (4a) [18,24]; while in other regions of its phase space it might also be periodic but with periodsT = pT where the numbers p are integers but might be very large, or it might even display an aperiodic, quite complicated (in some sense chaotic) behavior [25] (for recent progress in the understanding of this phenomenology see [26][27][28][29]). It then stands to reason that the solutions of the generalized model (14) with (14a) replaced by (25) (and of course x in (14b) replaced by z) shall again approach asymptotically solutionsincluding, from open regions of initial data, completely periodic ones -of the model (25) without F : entailing a remarkable, and quite rich, phenomenology.…”