Book Selection been better if Dr Rao had discussed these areas and optimal control theory more fully in a second volume. In spite of the book being produced with an outdated printing technology, the typography is visually pleasing, with only a small number of misprints. The binding of the reviewer's copy was beginning to disintegrate when he received it, but this is not a major complaint to make of a mathematical text of 700 pages costing only £2.95-this low price is partly achieved by a subsidy from the Government of India. In summary, this is a very good book in spite of its (few) flaws and, accordingly, it deserves a warm welcome in the student market. PAUL HUDSON

Optimisation and Operations Research (Proceedings, Bonn 1977) R.

HENN

et al. (Editors)

Springer, Berlin, 1978. 270 pp. $13.50 This volume comprises papers from the proceedings of a Workshop held in October 1977 at the University of Bonn. The format of this series of Lecture Notes allows impressively quick publication, using what seems to be the contributors' original maimscripts. Readers of the Journal may be somewhat surprised at the contents of the book, since there is little they would recognise as "Operational Research", and even the interpretation of "Optimisation" may seem stretched. Of the 27 papers, seven are in the general area of combinatories and complexity theory, two are related to inventory control, one is on queueing theory and the remainder range over problems of convex analysis, operator equations, measure theory and optimal control. The papers are analytically orientated, rather than algorithmic or applied, and will, in general, be of interest to the specialist reader. Among the more applicable papers, that of Baum and Trotter ("Integer Rounding and Polyhedral Decomposition for Totally Unimodular Systems") investigates conditions under which the optimal solution value of an integer programming problem is simply the round-up or round-down of the optimal solution value of the corresponding linear programming problem. This extends work previously done in the job scheduling area. A survey paper on the travelling salesman problem (TSP) by Grotschel and Padberg includes recent references to known "hard" problems, and these will doubtless become standard tests for algorithm writers. Given that many problems can be formulated as TSPs the state-of-the-art in this area is a reasonable measure of the progress in a small segment of combinatorial analysis. There seem few misprints in the book, and the layout of individual papers is generally excellent. The variety of typefaces is not disturbing, since few readers will refer to more than one paper at a time. lAIN BUCHANAN

Qualitative Analysis of Large Scale Dynamical Systems A. N.

MICHEL

and R. K.

MILLER

Academic Press, New York, San Francisco and London, 1977. xv + 289 pp., £16.00/$22.50 This book aims to present a mathematical theory that will have useful applications to the management and control of large scale dynamic systems, for example in power engineering, aerospace, industry, economics, society, and ecology. Most systems of this type have certain general properties in common, such as interconnection of subsystems and high dimensionality, that make it worth looking for a unified theoretical approach. Mathematically, these systems can be represented by ordinary differential equations,

499

Operational Research Society is collaborating with JSTOR to digitize, preserve, and extend access to Journal of the Operational Research Society. ® www.jstor.org