C.C. Baskin & J.M. Baskin (1998) Pp. xiv+666. Academic Press, San Diego. ISBN 0-12-080260-0. Price not supplied (hardback). Anyone with an interest in seeds will not need reminding of the Baskins; others need to know they have spent the last 30 years working their way through the dormancy-breaking and germination requirements of much of the herbaceous flora of the eastern United States, with occasional forays into shrubs and neighbouring regions. This book summarizes their efforts, and those of a staggering number of others. Writing it has taken nearly 10 years. In a preface they make the slightly unusual claim that they cite only those papers they have actually read. Clearly, they feel that this cannot be said of all citations by other authors, and perhaps they are right. Following a short introductory chapter, Chapter 2 is entitled ‘ecologically meaningful germination studies’. For those embarking on a career in seed biology (and perhaps for some who have already done so), this chapter alone is worth the purchase price. It is full of good advice. For example, resist the temptation to use artificial dormancy-breaking treatments, such as H2SO4. These may give you the satisfaction of germinating your seeds, but they will tell you little about what does this job in the field. The Baskins are also suspicious of green ‘safe lights’, which they regard as distinctly unsafe, and also of fancy statistics. In 1966 they asked the late Lea Barton how many seeds they should use in germination experiments. She told them that the statisticians at the Boyce Thompson Institute recommended three replicates of 50 seeds each, and the Baskins have followed this advice ever since. How long should a germination test last? Two weeks; if germination takes longer than this to start, there is always a suspicion that the germination conditions are influencing dormancy rather than germination per se. Finally, fungi are good at quality control; if your seeds are always attacked by fungi, you’re probably doing something wrong, or your seeds are dead. Subsequent chapters cover the types of seed dormancy, the ecology of seeds with different dormancy types, persistent seed banks and within-species variation in dormancy and germination, while three massive chapters survey the germination ecology of all the world's major biomes and of plants with unusual habitats or life cycles, e.g. parasites, halophytes and orchids. These chapters contain a wealth of fascinating detail. The Baskins take issue with the long-established term ‘induced dormancy’, since they know of no species with completely non-dormant seeds at maturity that has subsequently been induced into dormancy. No prize is offered for the first person to discover such a species, but research students should obviously be on the alert. The Baskins are rightly critical of the way much of the information on persistent seed banks has been collected and (worse still) reported. In many (most?) studies, it is impossible to tell if the seeds recovered are part of a transient or persistent seed bank. They also address the relationship between dormancy and longevity of seeds in the soil, and rightly conclude that there is none. Seeds that persist for years or decades in the soil may have physical (hard-seeded) dormancy, they may cycle continually between the dormant and non-dormant states, or they may be permanently non-dormant. The typical Baskin approach is pragmatic; they accept the widely-held belief that shade species in tropical forest generally have larger seeds that gap species, and choose to ignore the fact that seed size is strongly linked to phylogeny. Their epic surveys reveal many surprises and unexplored avenues. Why do temperate weed and non-weed floras have essentially the same spectra of dormancy-breaking requirements? We know a lot about the germination of parasites that happen to be pests, and almost nothing about the rest. We know far too little about how the dormancy and germination of aquatic plants is affected by low O2 conditions. A final chapter attempts to pull together some conclusions. The core of this chapter concerns the evolution of dormancy in the seed plants. It is clearly a great pity that we know so little about fossil seeds. Clearly, however, dormancy types are relatively ancient characteristics of families and orders. Dormancy is seen as a way of arranging for germination at a suitable time for establishment and survival. The most ancient form of dormancy, an underdeveloped embryo, is a relatively blunt instrument for achieving this. The evolution of the more advanced physiological dormancy has permitted much finer tuning and moreover it can easily be turned on and off. ‘Monumental’ is an overworked word, but it is one that applies to this book. The Baskins begin Chapter 1 with Theophrastus, the father of seed germination ecology, and conclude that he would be pleased with the progress of the subject. Much of his pleasure would be owing to the Baskins.