Abstract
It is pointed out that in the band theory of magnetism the magnons have frequencies comparable to the Fermi energy. Therefore, in the calculation of the magnon spectrum of iron, nickel, cobalt, etc., it is the time- or frequency-dependent response function of the electrons which is used, and this function—in contrast with the static response function—does not have a "Kohn kink" at q=2kF. But it is found that in the intermediate-coupling regime (exchange force in the range EF to 43EF) the magnons have a small range of momenta near q=kF where their velocities may be zero or negative, and it is shown how this part of the spectrum may be easily calculated, or approximated analytically.