When two states |k> and |k'> degenerate, a perturbation H' would lead to an energy split of <k|H'|k'>. As the number of degenrate states increases, the order of the secular equation rises correspondingly (and the equation could hardly be solved ?) My question is: is there any knowledge of the distribution of the energy levels of the "good" states (linear combination of the original degenerate states) when the number of degenerate states is big? Are they mostly in the range of E0±<k|H'|k'>? Or, is there any knowledge of their range of spread ? Thank you.