This book provides a state of the art summary on the general theme of descriptive multivariate analysis. It consists of a collection of commissioned, edited articles by an international group of leading researchers: Phipps Arabie (Rutgers University) writes on "Clustering from the Perspective of Combinatorial Data Analysis". Bernard Flury (Indiana University) includes some highly novel and previously unpublished work in "Principal Component Models for Patterned Covariance Matrices". David Edwards' (University of Novi Nordisk, Denmark) "Graphical Modelling" and the chapter on "Convergent Computation" provide thorough and useful surveys of recent research not available in other texts on multivariate analysis. Other contributions include James Ramsay (McGill University, Montreal); Clause Wiehs (CIBA-GEIGY, Bosle); Willen Heiser (University of Leiden); and Ruben Gabriel (University of Rochester, New York). The material should provide a useful reference for graduate students and researchers.