The Herefordshire ‘Lagerstätte’ is so-named because it represents one of those unusual fossil deposits that preserves not just the commonly found hard parts of animals, but the soft parts as well. The fossils of these small marine animals, approximately 425 million years old, are preserved as crystals of calcium carbonate within hard nodules composed of similar material. The fossils are not crushed, but retain their full three-dimensional form. The study of these fossils requires that we understand their morphology in as much detail as possible, but three-dimensional shape is difficult to interpret from the flat surfaces produced when the nodules are initially cracked open. It has not proved practical either to dissolve or dig the surrounding nodule material (‘matrix’) away from the fossils to reveal their three-dimensional form. Medical CT (or CAT) scanners have allowed us to see fossils inside the nodules, but these machines are expensive to use and are unlikely to be capable of resolving sufficient detail.
We describe a technique for overcoming these problems in which we repeatedly grind away small thicknesses of material from specimens and at each point photograph the surface with a digital camera. The resulting set of images represents the three-dimensional morphology, although the fossil is destroyed in the process. Computer video files are used to package these image sets for initial inspection. Three-dimensional computerized reconstructions can be created by a method involving the manual tracing of outlines on each image, but we consider this approach inappropriate in this instance. Datasets from grinding are treated instead as ‘volumes’—three-dimensional arrays of cubic ‘voxels’, which can be thought of as the three-dimensional equivalent of pixels. Volumes are ‘rendered’ (virtually photographed by the computer) either by a technique known as ‘volume rendering’, which produces images analogous to X-ray photographs, or by a two-stage process in which the computer first constructs a surface using the edges of the fossil as a guide (an ‘isosurface’), and then renders this by a realistic technique known as ray-tracing. Volumes are rendered from sequential angles, producing a sequence of images that are combined to produce a final video file of the fossil spinning on the spot. We illustrate the method with reconstructions of specimens from two familiar groups of fossils (a trilobite and a graptolite), and of a newly-described soft-bodied mollusc.
Reconstructions produced in this way can be edited (‘retouched’) easily and manipulated to remove blemishes (caused by variations and imperfections in the preservation of the fossil, or by small spots in the matrix of the nodule), to dissect out small structures for detailed study, or to combine both halves of the fossil (created when the nodule was initially split) into a single reconstruction.
The approach will provide the primary means of study of the Herefordshire fossils. It also has the potential to provide a means of obtaining or communicating morphological information in a variety of palaeontological applications.