Why do living organisms have the designs (and especially the skeletons) that they actually possess? Is it possible, and legitimate, to infer from the fossilised remains of a long-dead creature how it functioned as a living system, with all the components operating together in harmony? Some 40 years ago there was an often stated view that studies of functional morphology in fossil animals could never be more than clever speculation. Yet as time went by, it became increasingly clear that functional interpretations, when carried out in the right way, were indeed a proper field for study in palaeontology, and that animal skeletons, of almost any kind, could yield definitive information about how their bearers had lived. We need first to consider the origins of animal skeletons. There are two important factors here. The first is contingency, in other words the `accidents of history', which established suites of body plans which could subsequently be modified in different ways. Yet as ROGER THOMAS and WOLF-ERNST REIF pointed out in their `Skeleton-Space' model (1993), there are confining physico-chemical constratints which thereafter determine evolutionary pathways. There are, in fact, only a limited number of ways in which a skeleton can be functional, as determined by the properties of the material of which it is constructed, constraints upon growth and development, and the requirement for its component parts to function in terms of the whole organism. In consequence "the discovery of `good' designs ­ those that are viable and that can be constructed with available materials ­ was inevitable, and in principle predictable ... the recurring designs we observed are attractors, orderly and stable configurations of matter that must necessarily emerge in the course of evolution" (THOMAS & REIF 1993). Where then, with this in mind, do we proceed from here? Amongst compendia regarding form and function in fossils, we have the recent Functional Morphology of the Invertebrate Skeleton (1999), a fine collection of 43 papers edited by ENRICO SAVAZZI. Here one finds both specialised case histories and encompassing reviews, dealing with many kinds of invertebrate, and very useful it is regarding the various ways in which invertebrate palaeontologists study their fossils as living organisms. But the present volume is something different, for it encapsulates the refreshingly individual approach which has emerged in Germany over the last several years, most vigorously articulated by MICHAEL GUDO and his colleagues at the Senckeneberg Institute, Frankfurt am Main. Their basic concept is that the structural and functional constraints on living organisms can best be interpreted in terms of engineering analogues. Mechanical engineering, after all is about how machines are constructed and how they work, and there are simple analogues all around us. Consider, for a moment the evident correspondence between the claw of a crab and a pair of pincers, or an arthropod limb and the arm of a mechanical digger. There are surely many useful insights to be derived from an understanding of engineering principles, and the research papers collected in the present volume are a testament to the vigour of this approach. For herein we find not only concepts, but also tools and techniques in common use in engineering applied to biomechanics; computer-aided design and tomography, landmark analysis, Finite Element Analysis, and CAT-scans. Such tools give a much greater objectivity to analysis of function, for it is true enough, as Carpenter comments in this volume, that `theoretical models are often tainted with preconcieved ideas'. There are thirty papers in five sections, each of which consists of several papers, and at the beginning of each section is an explanatory introduction and summary. Section 1, Functional morphology and biomechanics. Following introductory comments by GUDO et al., there are six papers all concerned with vertebrates, and especially dinosaurs. To bring up a simple issue, how can we be sure about the stance of dinosaurs? We need to tackle this from the engineering point of view, and only if we do this can we be fully objective, and thus understand behavioural specialisations. The work presented here employs landmark analysis (BASZIO & WEBER), parametric modelling, allowing 3D digital reconstruction (the DinoMorph model) (STEVENS), and various other morphometric tools are developed and explained (EGI & WEISHAMPEL). Actualistic models, supplemented by CAT scans (CARPENTER) allow a precise understanding of how forelimbs are specialised in different groups of fossil predators, and Finite Element Analysis (FEA) is used to determine stress and strain in tyrannosaur bones, and so enables testing of ligamentary functions, and skull resistance to bending in other carnivores (SNIVELY & RUSSELL). Joint and movement functions (as the more elderly of us know too well) are adversely affected by continued stresses, and the skeleton must be designed to cope with this. How it does so is addressed by HENDERSON & WEISHAMPEL. The techniques now available have seldom been used before, and have a great analytical power for palaeontology. Section 2. Functional and Ecological Morphology, explores the interplay of biological, ecological, and taphonomic factors in reconstructing extinct organisms. Here again the emphasis is on vertebrates, both living and fossil. Some very interesting questions arise, and are answered. Was a bipedal posture possible for a Triassic lizard (RENESTO et al.)? Yes. Can limb reduction in living fossorial skinks be quantitatively correlated with burrowing performance (BENESCH & WITHERS)? Yes. What can be said about the three dimensional kinematics of climbing in various arboreal apes? Can this knowledge be applied to the understanding origins of human bipedalism (ISLER)? Probably. Is the optimisation of molar dentition in small rodents applicable also to larger animals (SCHMIDT-KITTLER)? No. Are structural specialisations of arvicoline rodent teeth related to diet (HERRMANN)? Yes. Are the teeth optimised in Miocene Hipparion for shea ring (KAISER)? Yes, they are, but without investment in unnecessary structure. Questions as these are fully explored in this section, and while the papers are on the whole less mathematical, they still rely directly on engineering principles. So do the three papers in this section dealing with invertebrates. Thus a modern coral living on soft substrates, swept by tidal action, can passively right itself if overturned, but only if its shape is precisely constrained (HUBMANN et al.). Both mathematical and physical models are used here to shed light on the orientative function of the nema in scandent graptolites (RANTELL & RIGBY). And encrusting bryozoans grow in different ways according to whether or not they are covering microbial mats; herein the different kinds of patterns are quantified (KASELOWSKY et al.). Section 3. Engineering and constructional morphology discusses how specific structures work in the context of the whole organism, and it is here that the use of FEA comes into its own. This means that structures can be modelled as a set of small, clearly defined elements interconnected at nodal points, and this has important applications. While some simplifications inevitably result, the models produced can be tested and evaluated independently. The use of this tool is here explained (FASTNACHT et al.), and in separate papers is applied to various problems. For example, the shapes of reptile and mammalian skulls relate, very precisely, to stress patterns generated by biting and supporting the head (PREUSCHOFT & WITZEL). The application of FEA to carnivorous synapsid reptiles (JENKINS) provides clear indications about comparative predatory habits, and hence niche partitioning. Likewise, the detailed analysis of the head of a dogfish (GUDO & HOMBERGER) provides evidence of mechanical constraints for the evolutionary history of fishes and tetrapods and bears very directly on the fish-tertrapod transition. In the final paper in this section GUTMANN combines a philosophical, constructional, and mathematical approach to problems of crustacean evolution He comments "Constructional morphology is a necessary prerequisite of evolution theory" Who could doubt it? Section 4. Constructional morphology and evolution applies the biomechanical approach to major issues in metazoan evolution, with particular reference to the constraints governing evolutionary pathways. OSCHMANN et al. set the scene with a consideration of the early evolution of the Earth and the origin of life. Here is stressed the importance of the ancestral `motiloids', fluid-filled membrane-bound protocells, from which came procaryotes and later the eucaryotes through endosymbiosis. While the ideas of GRASSHOF and GUDO have already been presented in the Senckenberg Poster The Evolution of Animals' (2001), here they are spelt out in greater detail. Moreover the concept of an ancestral `gallertoid' is here explained, as central to the development of compact bodies on the one hand and elongated bodies on the other. The nearest living creature to the hypothetical gallertoid is the placozoan Trichoplax, which is here considered (SYED & SCHIERWATER) and its evolutionary history reconstructed. Lastly an engaging and extended article by GUDO treats the evolution of the chordate Bauplan and its modifications. While some of the theoretic models here are going to be controversial, they are surely worthy of attention, and they should be much debated. These evolutionary pathways are after all dictated by structural and functional constraints, and their elucidation is founded on an understanding of engineering and biomechanics ­ what will, and what will not work. Section 5, Theoretical concepts. ethodology and philosophical analysis concerns the validity and methological justification of our current evolutionary framework. The function of the philospher, as ISIAH BERLIN pointed out many years ago is to question, maybe not too often, and maybe not too much, the structures and frameworks that govern our thinking and that we take for granted in our society. So, is a rigidly prescriptive cladistic framework really the best way forward for phylogenetic reconstruction? REIF provides another way of thinking, in terms of `reciprocal illumination' of pattern and process. He notes "The final goal of all systematic and evolutionary studies are integrated stories of the adaptive, synecological, and biogeographic histories of taxa. It does not seem prudent to stop short and restrict oneself to the construction of cladograms." Well said indeed! What of the geosphere-biosphere system? Is life truly a planetary phenomenon (LEVIT & SCHOLZ) ­ is the biosphere itself a living system of the highest order, part of an uninterrupted global morphospace? With such thoughts we end this remarkable collection of papers, so remarkably broad-ranging, but so germane to the main theme. I hope that there will be further papers of similar kind forthcoming in Senckenbergiana Lethaea. We have here truly a feast for the intellect. This compendium, for which we thank MICHAEL GUDO and his colleagues, will surely have a singular influence. I personally have learned immeasurably from it. I trust that it will stimulate other readers in the same way.