What Did Dinosaur Skin Look Like? Impressions Left in Mud

Soft tissue decays rapidly and is almost never preserved directly, so the usual evidence is an impression, meaning the pattern of the skin surface pressed into fine sediment that then hardened, leaving a negative of the texture. That requires a carcass to come to rest on or be covered by sediment fine enough to take the detail and to be buried before decay destroys the surface, which is a narrow window and explains why such specimens are rare. A smaller number preserve actual mineralised tissue, where minerals replaced the skin structure rather than merely recording its outline, and those are the specimens that supply information about internal structure, colour and thickness rather than only surface pattern.

The surfaces recorded are varied and none of them resembles reptile skin exactly:

• •Small non-overlapping scales, generally polygonal, quite unlike the overlapping scales of a snake
• •Larger feature scales scattered among smaller ones in defined patterns
• •Tubercles and bumps of various sizes across the body
• •Bony plates embedded in the skin in armoured groups, which fossilise directly
• •Filaments and feathers in numerous groups, preserved in fine-grained deposits
• •Variation across the body, with belly, back and limbs differing in the same animal

Discoveries from the 1990s onward transformed the picture. Specimens from fine-grained Chinese deposits preserved filaments and then unambiguous feathers on numerous dinosaurs outside the bird lineage, establishing that such structures were widespread among the predatory groups rather than confined to birds. Further finds pushed the distribution wider, and filament-like structures in groups on the other main branch of the family tree raised the possibility that the feature is ancestral to dinosaurs generally, which remains debated since those structures are less clearly feather-like. The practical effect is that reconstructions changed, with many familiar animals now shown feathered, and the argument has shifted from whether feathers existed to which animals had them and where on the body.

A small number of fossils go well beyond surface texture and they are correspondingly valuable. Specimens described as mummified preserve skin over much of the body in three dimensions rather than as a flat impression, which shows how the tissue draped over the underlying anatomy and gives body outlines that skeletons alone do not. Several such specimens of duck-billed dinosaurs have been recovered from North America, one of which preserves what appears to be a fleshy crest that the bones did not indicate. An armoured dinosaur recovered from a Canadian mine in 2011 preserves the animal essentially intact from the front half, with armour, scales and traces of pigment in place, and is among the best-preserved dinosaur specimens known.

The revision has been overstated in popular accounts, and the counter-evidence matters. Extensive skin impressions from large horned and duck-billed dinosaurs show scales across the body with no sign of filaments, which is direct evidence rather than absence of evidence, and comparable impressions from several large predatory species show scaled skin in the regions preserved. Body size is relevant, since large animals shed heat poorly and an insulating covering is a liability, which is why large mammals in warm climates have sparse hair. The current picture is therefore mixed, with feathers or filaments in many groups and particularly in smaller species, scales in others, and both on the same animal in different places, which is exactly what living birds show.

Skin is known mostly from impressions pressed into fine sediment rather than from preserved tissue, which needs rapid burial in the right material. Impressions show small non-overlapping scales unlike a snake's, with larger feature scales among them. Feathers turned out to be widespread among predatory groups, while large horned and duck-billed species show scales directly, so the picture is mixed rather than uniform.