In the nineteenth century, illustrators of palaeontological books let their imagination free and portrayed extinct reptiles similar to the more familiar mammals. Since then, these early illustrations have been used by paleontologists as caricatures, examples of unfounded reconstruction. The truth is that life style is at best difficult to reconstruct from the anatomy. Burrows are a physical record of behaviour, and so the study of their fossils allows to reconstruct the ecology without too much recourse to the imagination.
The habits of some burrowing therapsid reptiles could be recognized from their skeletons (Smith 1987). Adaptations for life in the ground are almost as striking as the adaptations to flight and to swimming. The very distantly related mole, golden mole, and marsupial mole all display these adaptations. So do two known dicynodonts from South Africa. Until the discovery and description of fossil burrows, however, it was not known that several other therapsid reptiles burrowed. These animals had no obvious adaptations in the skeleton, yet some likely lived the whole year in their burrow. The discovery of these burrows pushed the earliest record of burrowing in reptiles into the Permian (Smith 1987).
