Abstract
The Knight shift in beryllium has previously been measured to be small and negative, when referred to an aqueous solution of BeCl2. Theoretical calculations assume a reference consisting of a bare nucleus shielded by the core electrons. With the use of a recent measurement of the shielded nuclear magnetic moment in free Be+ ions and published Hartree-Fock wave functions of Be and Be+, it is shown that 20(4) ppm should be added to the experimental shifts in order to compare them with theory. This correction is of about the same magnitude as the measured Knight shift.