What Is a Skin Impression? The Outside of an Animal Pressed Into Rock

The impression is a mould rather than the skin itself, produced when a carcass settles into fine sediment and the sediment hardens around it before the tissue decomposes and collapses. That requires a narrow set of conditions: sediment fine enough to take a detailed impression, rapid burial, and chemistry that lithifies the sediment quickly, frequently involving microbial mats that stabilise the surface. Some specimens preserve the skin itself as a carbonaceous film or as a mineral replacement rather than only as a print, and those carry more information since the internal structure and occasionally the pigment can be examined. Preservation is patchy even in exceptional specimens, so a fossil may record skin over the tail and nothing elsewhere, which produces the awkward situation where a reconstruction is well founded in one region and guesswork in another. The distinction between an impression and preserved tissue matters for what analyses are possible, since a print supports geometry alone.

Accumulated impressions have settled several questions about appearance:

• •Scale patterns across many dinosaur groups, generally small non-overlapping scales rather than the large overlapping ones reptiles were assumed to have
• •Larger feature scales set among smaller ones in several species, forming rosettes and patterns that would have been visible in life
• •Feathers and filaments in many theropods and in some ornithischians, which is the finding that most changed reconstructions
• •Mixed coverings in single animals, with scales on some regions and filaments on others, which complicates any simple statement about a group
• •Pigment structures in preserved feathers and skin, which have supported colour reconstructions including countershading in at least one armoured dinosaur
• •Soft tissue outlines including crests, wattles and fleshy structures that leave no bone and that alter an animal's silhouette substantially

Even with impressions, reconstruction involves a great deal of inference and the honest practice acknowledges it. Coverage is partial, so an artist works from patches and interpolates. Scales and feathers are not distributed uniformly in living animals, so extrapolating from one region to the whole body is unreliable. Colour is almost entirely unknown except where pigment structures survive, and even then the reconstruction depends on interpreting those structures by comparison with living animals. Soft tissue that leaves no trace, including lips, cheeks, fat, air sacs and display structures, has substantial effects on outline, and a reconstruction that renders only what bone and skin document produces an animal that looks shrink-wrapped and is almost certainly wrong. That last point has become a recognised criticism within palaeoart, with illustrators arguing that reconstructions should include the quantity of soft tissue living animals actually carry rather than only what the fossils prove.

Skin impressions are trace fossils rather than body fossils, since they record a structure without preserving it, and that places them in the same category as footprints and burrows with the same interpretive cautions. They can be displaced, so an impression need not be adjacent to the animal that made it. They can be produced by processes other than skin contact, and features resembling scale patterns arise from sediment structures, which is why identification requires care. They are frequently found detached and attributed to a group rather than a species. Where an impression is associated directly with a skeleton the attribution is secure, and those specimens are correspondingly valuable. The broader point is that soft tissue evidence has expanded enormously as a category over recent decades, with feathers, skin, pigment, muscle outlines, gut contents and even preserved cartilage reported, which has moved the reconstruction of extinct animals from an exercise in guesswork towards one constrained by evidence.

The print is a mould left when fine sediment hardened around a carcass before the skin decayed, so it records geometry rather than tissue. Impressions show small non-overlapping scales rather than the overlapping reptile pattern assumed earlier, and feature scales arranged in visible patterns. Reconstructions that render only what is documented produce shrink-wrapped animals, which living animals never resemble.