A systematic study of the Lz 3M& (3s-2p) transition in elements from Cl to Cr reveals a satellite of unusually large intensity ( -30% of parent line). Previous work on some of these elements led to the suggestion that the satellite is due to a two-electron shakeup (configuration-interaction) process.We compare our measurements of position, width, and intensity of the satellite as a function of atomic number with the predictions of this model and extend the theory to explain the apparent disappearance of the satellite outside this range.