Abstract
The Palisade diabase sill of New Jersey has been widely cited as a type case of gravitational differentiation, and the present investigation is an attempt to explain certain inconsistencies which appear in previous accounts and to make them completely convincing. New data include 15 chemical analyses and micrometric measurements of 150 rock sections. They show that crystal fractionation can account satisfactorily for nearly all the differentiation phenomena displayed by the sill.