Abstract
Two independent quark-model predictions of the Λ magnetic moments, in agreement with one another and with experiment to 1%, are obtained by using two different inputs for SU(3) breaking in quark masses: (1) (msmu)=(MΣ*MΣ)(MΔMN); and (2) msmu=MΛMN. I discuss the implications of the surprising sucess of a description of quark magnetic moments as Dirac moments with quark masses related so simply to hadron masses.