The non-biomineralized fauna of the Herefordshire Lagerstätte (Briggs et al. 1996) represents a rare opportunity to sample the true taxonomic composition and palaeoecology of a Silurian benthic community. Fossils occur in spherical to subelliptical calcareous nodules found in a locally thick bentonite deposit in the Coalbrookdale Formation, Wenlock Series. They vary in size from under a millimetre to several centimetres, with worms, sponges, and non-biomineralized arthropods dominating the assemblage. All taxa are preserved in three dimensions, with no indication of post-mortem disturbance or compression. While some fossils preserve organic material (e.g., the graptolites) and others siliceous skeletons (e.g., the radiolarians), for the most part specimens consist of dark-coloured sparry calcite shapes within a lighter-coloured matrix (Briggs et al. 1996; see also Orr et al. 2000). Few unambiguous internal features of the organisms are known, but external details are commonly reproduced with a degree of fidelity sufficient to preserve fine spines 30 µm or less in diameter (Figure 1).
All taxonomic and palaeoecological study of the Herefordshire biota must be underpinned by careful resolution of the morphologies of the fossils preserved. However, while three-dimensional preservation is clearly richer in morphological information than the two-dimensional or near two-dimensional modes of most other Konservat-Lagerstätten (e.g., the Burgess Shale, see Allison and Briggs 1991), it introduces special problems. Surfaces exposed by splitting nodules are essentially randomly oriented with respect to the organism and show only those structures that happen to be intersected by the split. Structures of fossils are thus easily overlooked or misinterpreted. Furthermore, it is frequently difficult to distinguish the genuine termination of a structure, such as the body of a worm, from the situation where it passes out of the plane of section. These factors make attempts to reconstruct morphology from such ‘natural’ splits difficult and feasible only when a large number of specimens are known or when the morphology is relatively simple. If essential morphological data are to be obtained for the majority of taxa, to which neither of these conditions apply, then as much of the three-dimensional form as possible must be extracted from a single specimen.
The simplest and most common approach to determining the three-dimensional form of a fossil is to separate it from the matrix, either chemically or mechanically. Chemical dissolution techniques, though widely used in many branches of palaeontology (including the recovery of Konservat-Lagerstätte fossils such as the Cambrian ‘Orsten’ faunas; see Müller et al. 1995) require fossils and matrix to be chemically dissimilar. Although dissolution has been used to extract microfossils such as radiolarians from the Herefordshire nodules (Briggs et al. 1996), the predominantly calcitic nature of both matrix and macrofossils precludes its use in the study of most of the faunal elements. Mechanical excavation of the Herefordshire fossils is possible and has some utility for the study of large and robust structures (e.g., Briggs et al. 1996, fig. 1c, e). However, it is both extremely time consuming and inappropriate for resolving small or fragile structures, and thus is not a satisfactory approach for the study of the fauna as a whole. In view of these difficulties, our study has concentrated not on the physical isolation of the fossils but on techniques based on serial slices, in which many cross-sections are obtained from parallel planes intersecting the fossil. Data of this form can be collected in several different ways, but all approaches produce datasets that can be reconstructed into three dimensions with the aid of a computer.