Ernst Mayr is a venerable figure in evolutionary biology and this book is the result of his attempt to write a "life-history" of biology that would introduce the reader to the richness of biology and his passion for it. He has always seen biology as a quite different kind of science from the physical sciences. In biology, he argues, concepts play a far greater role than laws. He rightly dismisses the relevance to the history of biology of Thomas Kuhn's ideas - mainly derived from physics - on scientific revolutions and incommensurability. When thinking about the nature of explanation in biology he found the philosophers of science no help, and I am pleased to see he includes Karl Popper, who is indeed much overrated.

He has specifically omitted molecular biology, genetics and the neural sciences, and thus the title is rather misleading, for the book covers only a fraction of biology, almost all of it closely linked to evolution. He has left out molecular biology, which many would regard as the triumph of modern biology, partly because he feels that he is not competent to deal with it, but also because he thinks molecular biologists have discovered many trees but have not yet seen the forest.

This view of molecular biology is probably a reflection of his antireductionist position. He claims that while all biological processes are compatible with the laws of physics, the physical sciences cannot address many aspects of living nature and living organisms cannot be reduced to mere physicochemical laws. But he does not fully explore this somewhat dubious position and I, by contrast, maintain that all the interesting advances in biology have a strong reductionist element - molecular biology being the outstanding example. There is nevertheless an interesting problem in finding the right level of reduction, and when not to bother going lower down. This is particularly true of the behavioural sciences.

For Mayr, the essence of biology involves an historical narrative with evolutionary schema being paramount. Thus physiology, animal behaviour, genetics, cell and molecular biology are barely to be seen, even though he includes them in his list when discussing the structure of biology. While there is much truth in T. H. Dobzhansky's statement that nothing in biology makes sense except in terms of evolution, most of the research in the field is driven by quite different concepts. One can think about how the embryo develops or how a muscle contracts with only minimum reference to evolution.

̽»¨ÊÓƵ

He sees the functioning of the nervous system and embryonic development as the two remaining great problems in biology. Unfortunately his appreciation of the developmental biology and its history is rather unsatisfactory. For example, he wrongly suggests that M. Malpighi followed in Aristotle's epigenetic footsteps; quite the contrary - he was a preformationist. His description of modern developmental biology is, alas, very old fashioned and gives no sense of the enormous progress in our understanding over the past few years. The principles of development are pretty well understood, although there is an enormous amount of detail to be filled in.

Developmental biologists also have a better appreciation of the relationship between development and evolution with which he credits them. For example, the presence of gill-like slits in our own development reflects their presence in the embryos of our ancient ancestors not the ancestors themselves. The similarity in the mechanism controlling the development of our arms and those for the fly's wings are not only very satisfying but of great evolutionary significance, for they emphasise the striking conservation of developmental mechanisms in a variety of organisms. Evolution seems to have been quite lazy in respect of developmental mechanisms - when it found a good one it used it time and again.

̽»¨ÊÓƵ

Mayr is at his best when dealing with evolution itself as well as biodiversity. His summary of the early history of evolution is excellent, particularly of Darwin's monumental contribution. His analysis of the concept of speciation, a key feature of evolution, is excellent and he has contributed much to this area. Taxonomy is of great importance because it provides a picture of existing organic diversity and is essential for reconstructing evolutionary history. He provides a useful, if very critical, discussion of W. Hennig's system of cladistic analysis, which is based on clearly shared derived characters, while ancestral characters should be ignored. But even with evolution the selection of topics is a bit puzzling; there is, for example, no reference to J. Maynard Smith's discovery of evolutionarily stable strategies, which make sense of a very wide set of animal behaviour in evolutionary terms - for example, why in mating competitions within a species all the animals become neither "doves" nor "hawks".

The most interesting chapter raises questions about the relationship between ethics and evolution. Before Darwin it was conventionally believed that morality came from a Christian God. But evolutionary theory changed all that, for his theory of common descent deprived humans of their special place in nature. There was no longer a fundamental difference between humans and animals. In addition, the theory of evolution by natural selection got rid of any supernatural forces and of the idea that the natural world had been designed or that it had a divine purpose.

But could not our understanding of evolution provide new insights into the nature of morality and perhaps even provide a basis for ethics? This does, however, go against the view, to which I subscribe, that scientific understanding is value-free and cannot provide any basis for an ethical system. Nevertheless Mayr suggests that "a genuinely biological ethics which takes human cultural evolution as well as the human genetic program into consideration would be far more consistent internally than ethical systems that ignore these factors". At the very least, new scientific understanding of, for example, the environment, raises serious ethical issues that cannot be avoided.

There is ample evidence for altruism in humans and at first sight this seems at odds with natural selection, for does not selection always favour those individuals who are completely selfish? In terms of reproductive success, altruism benefits another organism at the expense of the altruist. But the partial solution to this apparent paradox is to think in terms of selfish genes, a concept to which Mayr does not refer. As J. B. S. Haldane once said, he would lay down his life for the survival of eight of his cousins, for that would be a good way to ensure the survival of his own genes. The modern concept of the contribution of an animal towards the survival of its relatives is due to W. D. Hamilton and is known as kin selection.

̽»¨ÊÓƵ

Pure altruism that gives no advantage to the altruist may be a purely human characteristic and may have contributed to the well-being of the cultural group. Ethical behaviour may be adaptive and so favour group selection. He is thus sympathetic to C. H. Waddington's suggestion that ethical learning is special, with a biological basis of a type akin to imprinting. If this is true, then early ethical education is of the utmost importance. An important aspect could be the need to learn to cooperate within a group and to recognise the advantages of such cooperation.

Mayr is suspicious of traditional ethical norms in the West - they are too rigid and also they have failed to take account of the enormous changes in our society. He dismisses abortion as murder and sees excessive egocentricity as a major ethical problem. But for him our greatest ethical problem lies in our responsibility for nature as a whole.

Lewis Wolpert is professor of biology as applied to medicine, University College London.