Equivalence between the Edwards-Anderson and Luttinger models of a spin glass

Abstract

The random “bond” model of a spin glass, as proposed by Edwards and Anderson, is demonstrated to be a special case of the random “Site” model advanced by Luttinger. An explicit relationship between the seemingly different order parameters employed in the two models is established, and for the pure spin glass phase (i.e. no ferromagnetic transition), the two order parameters are shown to be identical.