# shouldn't we add the oscillating terms into Bohr-Sommerfeld quantization formula

shouldn't be the quantization formula (in one dimension) equal to

$N_{smooth}(E)+N_{osc}(E) = \oint_{C}p.dq$ ??

where the Oscillating term is just the correction from Gutzwiller trace formula or a sum over Orbits

why is just the oscillating term ignored ..

-
can you justify? The traditional higher order corrections are simply fractions of integers on the left hand side which correspond to reflection phases for reflecting orbits. – Ron Maimon Apr 8 '12 at 20:27
i was more interested in the inverse problem ¨can we fromthe smooth and oscillating part of the eigenvalue staircase obtaine the potential $N(_{osc}(E)+N_{smooth}(E)=A \int_{0}^{a(E)} dx (E-V(x)^{1/2}$ in this case the inverse of the potential$V^{-1}(x)$ should depend ¿even in the semiclassical approximation on the SMOOTH and OSCILLATING part of the eingenvalue counting function – Jose Javier Garcia Apr 8 '12 at 21:17
： Then you are asking about the problem of reconstructing the potential from the energy levels, which is impossible without symmetry conditions. This was asked here before... – Ron Maimon Apr 8 '12 at 22:43