No one was around to write about dinosaurs or paint their images in caves. No frozen corpses have survived, as is the case with the relatively modern woolly mammoths. Every scrap of dinosaur knowledge comes from the study of fossils. So what are they?
Briefly, a fossil is any relic or representation, such as a bone or a footprint, that has become petrified - turned into stone - by the action of pressure and millions of years' exposure to dissolved minerals. A fossil skeleton is actually a stone skeleton - the bone has gone, to be replaced by stone. Most of the animal fossils we find are of sea creatures - they die, sink and are covered with sediment. But land animal fossils are rare - their bones would have been scattered or eaten. Only if the animal fell into a swamp, for example, or was quickly covered by sand, would a good skeleton survive.
So each dinosaur skeleton we have is the result of two National Lottery-type chances - that the creature died in circumstances that caused its skeleton to survive, and that it was discovered and recognised in modern times, perhaps in a quarry or a mine, or exposed by erosion. Our knowledge of Tyrannosaus rex, for example, is based on only three complete skeletons and a couple of dozen incomplete ones. Many types of dinosaurs probably remain to be discovered.
Footprints in the sand
Important for understanding social behaviour and the mechanics of movement, are "trackways" - sets of dinosaur footprints, preserved by being filled in with sand and soil. Thousands have been found, and some tell a story - one set in Australia shows how a predator dinosaur tore into a group of 150 small dinosaurs beside a stream, causing a stampede.
Evidence of deduction
Although we have only the bones and the footprints, comparisons with modern animals take our knowledge a long way. We know what the teeth of plant-eating or meat-eating animals look like. We know how bones are constructed to take the strain of tendons and muscles. We can judge whether a leg is built for speed or not. The degree of scholarship and skill brought to this kind of work is remarkable and exciting.
Importantly, too, it is not just deployed on new discoveries. A crucial aspect of dinosaur study involves revisiting specimens time and again, using fresh knowledge as it develops. So the vision of how a particular dinosaur looked and behaved can change over the years. For instance, when Tyrannosaurus rex was discovered in 1900, it was deemed too big to move fast. So it was drawn looking comparatively puny. But study of its fossilised leg and ankle suggested it probably could move quickly, and later pictures show far more muscle. Jurassic Park has it chasing a car. But such speed seems unlikely - had it tripped, it would probably have killed itself.
The model of an iguanodon in Crystal Palace Park, south London, made in 1854, shows a heavily-built animal standing on four legs, with a short horn on its nose. But the discovery of 30 complete skeletons in Belgium in 1878 revealed that the "horn" was actually a spiked thumb, and that it moved on two legs.
Palaentologists warn against too much speculation on what the dinosaurs were like. John Martin, a palaeontologist at Leicester City Museum, says:
"Much of what's published will be hypotheses based on what's possible rather than on anatomy." There are important science lessons here for children and their teachers. They can learn, for instance, how easy it is to build up layers of conjecture. ("The teeth were that shape, so it must have bitten its food that way - so it must have moved its neck like this, so the neck muscles must have . . ."). Artists' reconstructions, showing a particular colour, texture, stance, or method of eating, can all too easily be accepted as factual.
"How do they know?" should be a constant query. The whole area is a powerful demonstration of the power and limitations of scientific enquiry.
In cold blood?
Academics puzzled over whether dinosaurs were warm-blooded or cold-blooded for years. Today's lizards and snakes are ectothermic (cold blooded). They have no internal mechanism for generating body heat. Instead they pick up heat from their surroundings. Mammals, by contrast, are endothermic (warm-blooded). They use food as energy to create body heat. Dinosaurs were reptiles, so it seems fair to assume they were cold-blooded. But doubts have always existed about cold-blooded creatures' ability to maintain an active, predatory lifestyle. So the question has important implications for our understanding of dinosaur diet and behaviour.
Academics over the years have believed passionately in one interpretation or the other. Now though, the debate is less heated. Most recent research reveals such a variety of dinosaur types and lifestyles that one simple answer is not acceptable. Dr Angela Milner, an expert based in London's Natural History Museum, says: "They were quite unlike any animal living today."
Dinosaurs in disguise
Many experts have long believed birds evolved from meat-eating dinosaurs. The recent discovery in China of the remains of small meat-eating dinosaurs with simple feathers greatly reinforces this idea.
Process of elimination
The sudden extinction of the dinosaurs has long puzzled scientists. Many theories exist - an asteroid collision that caused the earth to be covered with a pall of debris that blocked off the sun for many years is a favourite. Layers of rock certainly seem to show a sharp global discontinuity that might be explained by a cataclysmic event. But however it happened, the removal of the dinosaurs cleared the way for the rise of the mammals and, eventually, of Homo sapiens.