T. rex Was an Orca on Land. Almost Nothing Else Matches the Movies.

Lex Fridman starts with Tyrannosaurus rex, perhaps the most famous predator in the history of life, and asks paleontologist Dave Hone what you would actually see if you stood in front of one. Hone, who lectures at Queen Mary University of London, co-hosts the Terrible Lizards podcast, and writes on dinosaur behavior and ecology, says the first impression is almost trite because everyone already knows T. rex is huge — until you stand there. Downstairs from his office he keeps a cast of a Smithsonian skull. He can fit comfortably through its mouth. It would be just about capable of swallowing him whole, and he is a big man. It is not even a particularly large T. rex. They get bigger than that.

The numbers are easier to say than to feel. Hone puts a typical adult at about twelve meters long — roughly fourteen yards — four and a half or five meters to the top of the head, and around seven metric tons, or about eight and a half short tons. His colleague Tom Holtz described the animal as an orca on land: a killer-whale-sized predator that happens to stand on legs. A very large African savanna elephant is five to six tons. T. rex is heavier than that, bipedal, and built only to eat meat. A big lion is about 200 kilograms, or 430 pounds. That is the gap Hone wants in your head before any conversation about hunting or intelligence.

People misread living giants for the same reason museum mounts flatten dinosaurs. Hone used to work in zoos, and he noticed that elephant houses, doors, bars, and food are all built to the animal's scale, so you never see the body against a human frame of reference. At London Zoo a rhino shoved its head through a gap in a wall and the head was twice the size he had expected; he once stood next to an elephant closer than Fridman is to him in the studio and thought, they are so much bigger than I thought. Skeletons do the same trick. There is space, a railing, a raised platform. When you actually stand under a T. rex, the foot finishes at the human knee.

Hone works through the animal with casts on the table: a boxy, robust skull, forward-facing eyes the size of tennis balls, and teeth that look larger than they are until you realize much of each tooth is root. He holds a T. rex tooth beside a cast from Carcharodontosaurus, a giant African carnivore that is not far off in length. In profile the teeth are a surprisingly similar shape and not far off in size. Turn them ninety degrees and the African tooth is about a third as thick. Thickness is strength. Combined with extra bone in the skull and extra muscle on that bone, T. rex has what Hone calls an uber-powerful bite and the ability to chomp through basically anything it wants. Other giant carnivores such as Giganotosaurus and Mapusaurus from South America may be a meter longer, but in mass he puts them at perhaps two-thirds or three-quarters of T. rex. Length is a python's trick. Mass is the point.

The arms are the opposite of the head. Hone's T. rex arm cast is not from the biggest individual, and it is still not far off the size of his own arm — and he does not weigh seven tons. There are two fingers. The supposed third digit is an extra metacarpal that does not support a finger. The deltopectoral crest, the radius and ulna, the claws, and especially the ligament pits all say the same thing. A deep pit means a big ligament, which means fingers that can hold a writhing animal without dislocating. T. rex has probably the smallest ligamentous pits of any tyrannosaur. Put those claws on a seven-ton animal whose individual teeth are the size of entire fingers, and the arm stops looking like a weapon.

The hind limb is where the engineering is. Theropods walk like birds, three toes down and the metatarsals nearly vertical, which lengthens the stride. In T. rex and in ostrich-like animals such as Gallimimus, the middle metatarsal pinches into a flattened diamond locked by the bones on either side, an energy-saving spring that evolved independently in several groups. The real motor is the tail. Hone describes theropods, with a paleontologist's wince at the cartoon, as doing something functionally like a lizard on its hind legs: a massive block of muscle on the first third to half of the tail, two and a half or three meters long in T. rex, anchoring on the back of the femur. Contract that muscle and the animal goes forward. Mammals walk with the rump. This is a longer lever.

How fast that lever could move has been argued back and forth. The most detailed recent estimate Hone cites clocks an upper end around 25 miles an hour, or 40 kilometers an hour, and probably less. Big animals cover ground because the legs are about three meters long, not because they gallop. He does not think a T. rex ran in the biomechanical sense of both feet leaving the ground at once; the loading will not allow it. With a four- or five-meter stride, power walking is enough. Fridman is reminded that cinema itself began as a bet about gait: Eadweard Muybridge lined cameras along a track to settle whether a running horse ever has all four feet off the ground. One frame showed that it does. A moving photograph followed. Large-animal mechanics have always been worth a wager.

Would a T. rex even consider two primates as food? Hone is not sure. Nothing in that world looked like us, and animals that have never met a thing are often simply cautious. The more practical filter is size. Carnivores typically take prey much smaller than themselves, despite every documentary that shows T. rex on an adult Triceratops of equal mass, or lions on a buffalo. For a paper on Microraptor, the small gliding dinosaur from China, Hone tallied stomach contents and came out around 5 to 20 percent of the predator's mass — in line with living carnivores, and not far from the rare consumed bones we have for tyrannosaurs. Five percent of a seven-ton adult is a lot of animal, but two humans might still be beneath the effort unless a smaller T. rex, perhaps only a ton, is the one that finds you. The Jurassic Park freeze is no help either way. Hone calls it nonsense. T. rex has gigantic eyeballs. Absolute size, not the eyeball's share of the head, is what buys acuity, which is why an eagle with eyes not much larger than ours can see the way it does. Kent Stevens published a graphic that makes the point with tennis-ball, even baseball-scale eyes. Hone has looked for the paper Michael Crichton later implied in *The Lost World*; as far as he can tell it never existed.

The menu, then, is not "whatever it wanted" in the joke sense. Adult Triceratops weighed as much as T. rex, carried horns a meter to a meter and a half long, and may have been aggressive. Hadrosaurs such as Edmontosaurus and Parasaurolophus lacked armor and horns, but they were simply enormous. Grapple an animal your own size and you risk a smashed foot or a body slam; even if you win, a five-ton carcass will rot before you finish it. Lynx take oversized deer, and lions take buffalo in groups, but those are exceptions. Almost every large carnivorous dinosaur with stomach contents or a healed, tooth-broken bone in a victim points at juveniles. A young Triceratops can still weigh a ton and be the size of a rhino. It does not yet have the full horn, frill, or practice.

Hone does not present that as a cruel specialty of tyrannosaurs. It is, he says, the rule of thumb from fish and starfish and praying mantises through crocodiles to big cats. Juveniles are small and weak. They lack the adult weapons. They are naive about wind, sightlines, and what a predator even is. They forage inefficiently, so they must feed longer, often in worse patches because adults chase them off the good ones — including the patch at the forest edge where a T. rex might wait. There are always exceptions. It would be very weird, Hone says, if dinosaurs were not doing what everything else does, and the bite marks and stomach contents say they were.

Bite marks are only as good as the burial story underneath them. Hone is blunt that you cannot take fossils at face value once you care about behavior: a lot can happen between death and the quarry, and three skeletons in one hole need not have died as a group. He edited a paper in which debris, fish scales, and small bones suggested animals might have been washed into a channel rather than living together. On a Mongolian duckbill held in a Japanese museum, colleagues who read erosion better than he does argued that the bite marks sit on top of weathering. The animal had died, washed onto a sandbank, and sat in abrasive sand for days or weeks. Then something chewed it. That sequence can only be scavenging.

The skeleton is otherwise pristine — one of the best hadrosaurs Hone has seen from Mongolia — and every mark is on a single bone, the humerus. Deep circular punctures land at the ends, matching the thick side teeth. Along the deltopectoral crest, where the big muscles attach, there are close parallel scratches. T. rex has a row of tiny incisiform teeth at the front of the mouth, a fraction the size of the lateral fangs, with a weird flat back. Hone's image is an Oreo: take the top off, then scrape the cream with your teeth. The same mouth kills and dismembers with the side battery and feeds with the little front row. Healed bone around a broken tooth is the opposite signal: a live animal that got away. T. rex, on this evidence, both hunted and scavenged. Guessing the mix is harder. Obligate scavengers such as vultures have to be ultra-efficient long-distance travelers. Hone's guess is that T. rex was still primarily an active carnivore.

The hunt he reconstructs is wolf or hyena, not cheetah. The feet are efficient compared with the herbivores; the top speed is only an upper estimate; the prey that matters is already large and not that quick. A four-meter animal is also almost impossible to hide. Lions struggle in long grass; T. rex lived in relatively open country, with stands of trees rather than forest, from Alberta down toward New Mexico and perhaps Coahuila. Hone suspects adults were largely nocturnal, using huge eyes and a strong sense of smell to close distance, then walking the prey down. A one-ton juvenile Triceratops or hadrosaur might be faster in a sprint and still lose on endurance. No predator wins every chase. The combination of size, vision, smell, and efficiency is at least a coherent way to survive in the open. Tarbosaurus, a very close relative, played a similar role in Mongolia. T. rex itself probably occupied most of western North America.

Dinosaurs were a global fauna — including Antarctica near the end — and Hone has written a book chapter on how many species that implies, because every correction pushes the number both up and down. Named, reasonably valid species sit around 1,500 to 1,600, with 40 to 50 new names a year for at least a decade and no sign of slowing. India is still thinly explored; new beds are turning up in places such as Ecuador; Argentina and Australia still have more rock than excavation. The method has barely changed since the eighteenth century. Either you split lithographic limestone in working quarries, which is how feathered animals and Archaeopteryx appear, or you walk the right age of rock and look for a chip sticking out. Whole skeletons do not lie on the surface. Bone turns to rock, takes up minerals, and then sand, frost, and moisture shatter it. What you want is the bit that has just begun to weather free.

Hone's favorite near-miss was in northern China: a colleague spotted less than a centimeter of claw in a hillside. Searchers moved on; diggers followed the bone. A couple of days later they had a nearly complete skeleton, later named Linheraptor, a close relative of Velociraptor, something like 90 percent complete, missing the end of the tail, with legs a little damaged by water. That is the dream, because a skeleton already on the surface is a skeleton already lost. Getting one out still looks like the nineteenth century: Paraloid glue wicked into porous bone, a plaster-and-burlap jacket, sometimes a backhoe, sometimes days of pickaxes, sometimes a helicopter because protected land in the American West forbids wheeled vehicles. Stan, the T. rex that later sold for $31.8 million, was months of field work and more than 30,000 hours of lab preparation, about 70 percent complete by bulk. The Larson brothers at the Black Hills Institute held it for decades and sold casts so widely that Hone has seen Stan's skull in Oxford and in a shop window in Lyme Regis.

The price stunned him. Plenty of T. rex specimens had sold for a few million; ten would have sounded absurd; Stan went to Abu Dhabi's Department of Culture and Tourism. That record has since been beaten by Apex, a Stegosaurus Hone does not find especially more impressive than other good stegosaurs. Sophie, a young but extraordinarily complete Stegosaurus at London's Natural History Museum, sold for around £250,000 — perhaps $400,000 — a decade ago. A much less complete animal is now worth a hundred times as much. Hone does not pretend this is a science market. It is closer to art: a thing is worth what someone will pay, and T. rex carries a premium because it sits with Triceratops, Diplodocus, Brontosaurus, and Stegosaurus on the short list almost anyone can name.

Asked how he would spend $10 billion on dinosaurs, Hone would drop on the order of half a billion on the best museum he could build — research, he says, is pointless if you do not communicate it — and spend the rest buying fossil land. The Solnhofen quarries in Bavaria, commercial and inherited private plots, a Brazilian lagerstätte mined as a business: train the workers, pay a bonus for every find, and hand the specimens to museums so no lab has to bid against a foyer. He has joked that all science funding should go to paleontology, because black holes and panda genetics will still be there in a million years, while a hillside can erase the only giant twice the size of T. rex in a season of wind. It is a joke, he stresses. It is also a field whose evidence is literally disappearing. Drones, seismic "thumpers" out of *Jurassic Park*, and radioactivity in uranium-rich bone have all been tried. Porous bone takes on the density of the rock around it, which is exactly what scanners need to differ. For now the instrument is still a tilted human head.

After the tangents, Fridman brings Hone back to the skull he calls incredibly robust and overbuilt. Hone prefers a different wording: it is the right amount of bone for the stresses, but compared with anything that is not a very large tyrannosaur it looks excessive. A figure from his old blog puts T. rex in orange beside Giganotosaurus in red. The linear sizes are similar. The front of the T. rex snout is bone; the antorbital fenestra in Giganotosaurus is enormous, something like half the skull. Openings in the back of the head and in the lower jaw are larger, and the jaw is thinner side to side. Same-sized heads, wildly different bone. Tyrannosaurs also fuse the paired nasal bones into a single rigid bar along the snout, a trait that appears even in early, human-sized forms. Hone suspects it first evolved to resist the pull of those little front teeth at the tip of a long nose, then became the chassis for a crushing bite.

The dynasty is old. Tyrannosaurs appear in the Middle Jurassic, around 160 or 165 million years ago, and last until the extinction dated near 66.5 million years — a hundred million years, most of it poorly sampled because the right rocks are scarce. Early ones would be chest-high, two or three meters long including the tail, with small heads and long arms, already feathered, already carrying fused nasals and the special front teeth. They show up in Europe, China, Kyrgyzstan, Russia. By the Early Cretaceous, animals such as Yutyrannus reach six or seven meters and perhaps half a ton to a ton, with three definitively feathered specimens. Hone has worked on Cope's Rule, the tendency for lineages to get larger: you cannot get much smaller than a small start, you can always get bigger, and larger bodies hunt more, move more efficiently, and win contests. Extinctions then reset the clock by killing the giants. Late in the Cretaceous the family splits. Long-snouted alioramins look more like giant Velociraptor, fast and relatively light-biting. The tyrannosaurines — Albertosaurus, Gorgosaurus, Daspletosaurus, a Chinese form Hone named, Tarbosaurus, Tyrannosaurus — build broader skulls and fewer, larger teeth. Only three or four of those become the ultra-giants ten meters and up.

Fridman reaches for "apex predator." Hone dislikes the term. In ecology it usually means a predator of predators, a great white eating sea lions, not a giant that eats herbivores. His friend Darren Naish has shifted to "arch predator" for a dominant giant that is not eating other hunters. T. rex is strange even among giant tyrannosaurs because it is, ludicrously, the largest carnivore in its ecosystem by a gulf. Hold a Velociraptor skull: some contemporaneous carnivores were a bit bigger, not enormously so. Hone's analogy is going to Africa, noting the lions, and asking for the next predator — a weasel. You do not get that gap in living systems. Herbivores were huge. Other huge carnivores were not. A Velociraptor-sized meal would be lions hunting mice. On land, asked who could challenge T. rex, Hone says nothing reasonable. South American carcharodontosaurs are comparable in length and lighter, with cutting teeth and muscular, clawed arms, more like wolves or hunting dogs bleeding prey down. Money stays on the heavier animal with the crushing bite. Spinosaurus in *Jurassic Park III* is a later argument; honey badgers chase off lions by being too expensive to injure, which is not the same as winning.

Hone prefaces the movie talk: he knows it is fiction, he gets asked anyway, and documentaries take liberties too. He also knows *Jurassic Park* is, by a ludicrous degree, the most popular media that touches his field, so the distortions travel. Velociraptor in the films stands taller than six-foot men, with domed heads; the real animal is about thigh-high to Hone, a meter or so at the head and around two meters long, with a long, narrow snout. Even in 1993, he says, researchers were fairly confident dromaeosaurs had feathers. Cheetah speed, giant size, pack hunting, and super-intelligence are the rest of the sticker. He has written at length that evidence for pack hunting in any dinosaur is almost nonexistent, and that the problem is the same taphonomic caution he applies to bite marks.

The famous Montana association of Deinonychus — a larger, older cousin often confused with Velociraptor, including in the books — with a Tenontosaurus carcass is the usual exhibit. The reading is a group that brought the herbivore down. Hone's first counter is blunt: why did they all die there? Lions eat the wildebeest; they do not all die beside it, and a species that slaughtered half its pride at every kill would not last. Predator traps offer a cleaner pile-up. At La Brea, a herbivore stuck in tar screams, carnivores wade in, and the ratios go absurd: on the order of three mammoths and one ground sloth against something like a hundred dire wolves and forty Smilodon. Cleveland-Lloyd is another crowded quarry. Toxic algal blooms can stack bodies the same way without tar. Hone does not think the Deinonychus sites are a tar pit, but he thinks the "successful pack" reading is the last inference those bones should earn.

Trackways are not kinder. A much-cited set of tyrannosaur prints is, as he reconstructs it, four or five tracks total: two from one animal, two from a second, one from a third, going the same way at similar sizes. Solitary animals do that. A ranger in South African national parks has seen leopards, which are almost antisocial, hunt together as a rare event. A female in heat leaves a trail; males follow hours apart; three sets of prints share a path and no social life. Spotted hyenas, Crocuta crocuta, star in documentaries as a mob ripping wildebeest. The scientific literature, Hone says, is that they mostly hunt alone even though they live in complex clans. Multiple tyrannosaurs in one bone bed might mean they lived near one another. It does not mean they hunted as a pack. Close kinship does not decide it either. Lions group; leopards, tigers, and pumas do not. African hunting dogs group; maned wolves do not. Spotted hyenas are social; brown and striped hyenas and aardwolves are not.

What would convince him is ugly and maybe undiscoverable: a herbivore trackway with several carnivores whose prints overprint one another in both orders, so A steps on B and later B on A, which means they were there within minutes, not days, and seven or eight of them converge as the ground goes chaotic. Fridman calls the work Sherlock Holmes. Hone, a Holmes fan, notes that Conan Doyle already borrowed Baron Cuvier, the anatomist who argued extinction from mammoths and claimed an animal could be restored from a fragment. Holmes's line is that he can restore events from the smallest detail the way Cuvier restores a body from a bone. Hone thinks some dinosaurs were social, some even pack hunters. Which ones, he has very little idea. Birds of prey such as ground hornbills do hunt in groups; crocodilians cooperate; even some lizards show social behavior. The capacity is there. The identifications are not.

Social life and sex are where paleontology is most tempted to overclaim. Medullary bone is the cleanest sexing tool Hone has: while laying, birds and many reptiles grow a weird, highly vascular bone inside large elements such as the femur and humerus, a calcium bank for eggshell. Slice a big bone, grind it nearly transparent, and the texture can mark a female in laying season. It does not mark males, young females, females out of season, sick females, or females that have already laid. A handful of specimens still contain eggs. Pelvis shape, so useful in humans, does not travel well, because human infants have enormous heads. Horns and frills do not split the sexes the way antlers do in many deer: every ceratopsian Hone can name, some forty species, has some horn and a frill. You can infer display. You usually cannot assign a skeleton a sex.

He prefers the term socio-sexual selection, because sexual attraction and social intimidation are hard to pry apart in bone. Lionesses find darker manes sexier; other males find them more intimidating. Black swans grow curly wing feathers in both sexes; each prefers curlier partners, and curlier females also win fights over nest sites. The curl is an honest signal because those are primary flight feathers and they make flying harder. Pale-maned lions persist nearer the equator because a jet-black mane overheats. Hone reads the little crests and brow horns of tyrannosaurs, Ceratosaurus, Dilophosaurus, and allosaurs the same way: the sexiest look is also the one that gives you away to prey, which is the point. If I am this bright and this ornamented and I can still run you down, look how good I am. Herbivores in loud groups of twenty ten-ton Triceratops cannot hide anyway, so they grow the largest billboard they can.

When both sexes carry the same ornaments — every Triceratops, every T. rex as far as the fossils show — Hone hears a hint of mutual sexual selection, the pattern in many seabirds, starlings, and parrots, where males and females invest similarly in rearing and therefore both choose. That would imply cooperation in reproduction, a weak signal with a large story. The problem is the sample. He puts more than 90 percent of dinosaur species as known from a single specimen, often a bone or a tooth; the number of species with enough complete skeletons to compare is closer to five or six than to ten. A single spectacular animal is not a population. Microraptor, later in the conversation, is his cautionary tale in diet. The same caution applies to sex.

Hone's method, when he trusts a lifestyle, is independent lines of evidence that are not the same measurement twice. The failure mode is the opposite: a superlative specimen treated as a silver bullet. Microraptor, a crow-sized glider from China, now has at least a dozen good skeletons and several with gut contents. Hone has described one with a small mammal foot; others hold a bird, a lizard, a fish. At least two papers concluded that these were fish-eating animals. His translation: that one ate one fish once. Taken together, the simplest reading is a generalist caught eating different things at different times. It is also possible one individual had learned fish. A single meal cannot decide.

Spinosaurs are his positive case. The group is fragmentary and badly known, but the gestalt repeats. The head and teeth are crocodile-like compared with every other carnivorous dinosaur; mechanical analyses find a similar function. A spinosaur tooth is nearly circular in cross-section, distinctive, in the company of crocodiles, dolphins, and fish-eating fish. They turn up in or near water more than other carnivores from the same rocks. Isotopes in the teeth often correlate with crocodiles, fish, and turtles rather than with land dinosaurs living alongside them. Baryonyx, the British member of the group, was found with fish scales in its chest. Hone does not claim a fish-only diet — they took terrestrial dinosaurs and pterosaurs too — but he thinks it is very hard to argue that this was just another land hunter. He and Mark Witton have a book, *Spinosaur Tales: The Biology and Ecology of the Spinosaurus*, due in the UK in November and in North America in early 2026, with Witton as co-author and artist.

His ecological picture is a giant wader, a very weird stork or heron, not a pursuit diver. Papers that made Spinosaurus a strong swimmer are, in his view, weak; he and Tom Holtz wrote a blunt critique, and some of the original authors have since reversed themselves. The animal is extreme even inside an already strange family. The jaws wave, narrow, and open into a rosette; nostrils sit back from the tip; the neck is long; the back carries a sail of elongated spines; the tail, in Spinosaurus at least, flares into a newt-like paddle; the arms are large and clawed; the legs look short. Baryonyx in Britain and Suchomimus in Niger share the head, neck, and arms without the sail, the paddle tail, or the short legs. Length is not mass: a snout in a Milan collection implies perhaps 15 meters, longer than any T. rex, with a narrow chest rather than a barrel. In a fight he still takes T. rex. The Jurassic Park III animal snaps a tyrannosaur neck; Hone notes that Spinosaurus necks are strong up and down and weak in rotation, while T. rex has among the strongest necks known. Weakest twist against strongest neck, long thin jaws against a crushing skull. He does not buy it. He does, for its era, rather like the JP3 reconstruction.

Asked what the franchise gets right, Hone starts with unglamorous hits. T. rex is about the right size and shape, and in the Jeep chase it never truly runs: watch and one foot stays down. Accuracy otherwise lurches by sequel. He works on pterosaurs; the Pteranodons in *The Lost World* are a throwaway, the ones in *Jurassic Park III* he finds mostly excellent, and the *Jurassic World* animals a regression, including a dramatic snatch-and-carry that Pteranodon feet and lift could not do. A creature designer he will not name told him that fixing a model to a new brief is no more work than the constant tweaks directors already demand. Someone still added teeth to Gallimimus in a *Jurassic World* children's book — the original animal has a beak — on a beast that only sprints past the camera. After the first film, Spielberg or Universal funded the Jurassic Foundation, small grants Hone has seen help a PhD student. Later films, as the property became a franchise, show less of that engagement. He still tells people not to treat the movies as documentaries: nobody watches James Bond to learn espionage. A colleague at the Royal Tyrrell Museum in Alberta had visitors, more than once, angry that the real dinosaurs were not out back.

Intelligence is where a recent paper ran ahead of the skull. Dinosaurs, like most reptiles, do not pack a brain to the inner bone the way we do; an endocast captures brain plus surrounding tissue, and the olfactory and optic bulbs at least show how much of that volume went to smell and sight. CT of the bony inner ear can even suggest the frequencies an ear was built for, which Hone finds phenomenally cool and not very informative. Neuron density is the trap. Birds pack extra cells with unusual folding, so a small crow brain can be surprisingly smart. A controversial paper pushed T. rex toward primate-level neuron counts; within months a team of paleontologists and neurologists — Hone refers to it as the Casper paper and supports the revision — brought the estimate from three billion neurons down to a range of about 250 million to 1.7 billion, in crocodile territory. Mass estimates and brain-volume estimates can each be off by 20 or 30 percent. Make the brain large and the body light and you have a genius; reverse it and you have a dullard. Crocodiles are smarter than folklore allows. They are not monkeys. A T. rex brain is a fraction of a chimpanzee's, in an animal that weighs seven tons.

Cannibalism, by contrast, is close to certain, and it is certain in part because T. rex was the only large carnivore in its world. A tyrannosaur jaw from Alberta has a T. rex tooth embedded, with overgrowth; a T. rex foot bone carries the distinctive feeding scrapes. Anywhere else, several big hunters could have left those marks. In a T. rex ecosystem, a large chewed bone has one author. Hone separates that from the facial brawling written on tyrannosaur skulls: half to two-thirds of adults show scarring and healed injuries. The foot does not. It is unhealed, bitten more than once, with feeding traces rather than puncture-and-run combat wounds. Nobody fights by repeatedly scraping someone else's foot. Eating other carnivores, especially your own species, is still rare, and the reason is parasites. Predators accumulate them; another lion is the meal most likely to carry the suite that infects you. If you are starving, or too maimed to hunt, and two tons of meat is on the ground, the calculation changes. Hone ends his behavior book on that tone: he has spent chapters knocking down overconfident stories, then notes that new techniques, new data, and simply more specimens of known species can still build something stronger. Growth, dimorphism, and population structure do not come from one or two animals. They come from 50 or 100.

The end, Hone says, is the story most people now know, and it was probably fast. About 66 million years ago an extraterrestrial body hit just off the Yucatán Peninsula, atomizing itself and, worse, the particular rock it struck, throwing up an ash cloud he compares to a nearly instantaneous nuclear winter. He cites a description of a piece of rock about the size of Mount Everest traveling at something like ten times the speed of sound. Local vaporization is obvious. The lethal part is climate moving faster than life can track. Earth's five recognized mass extinctions are, in his telling, all forms of climate shock — volcanoes, oxygenation, ice. Knock out a moth that pollinates a tree whose nuts feed a squirrel that feeds a cat, and four other things go with the moth. Big terrestrial animals are the worst combination: they need more food and more land, breed slowly, hold less genetic diversity, and cannot swim to another hemisphere when a desert or a mountain range stops them. Water buffers heat. Whales can leave. Dinosaurs were mostly large and mostly on land. It is not surprising they did badly. Paleontology's practical use, he adds, is a preview of who suffers first under extreme climate stress.

He would be amazed if no non-avian dinosaurs lingered in some New Zealand or Hawaii of the earliest Paleogene, for a few hundred thousand years or even a couple of million. Reports of Cretaceous dinosaurs above the boundary have not held up; prairie dogs and roots move bones. On a global map those refuges would be dots, and the odds of finding, dating, and recognizing them are almost nil. That is ordinary biology: lemurs on Madagascar, New Zealand's oddities, Galápagos tortoises, Australian marsupials. Mammals, crocodiles, birds, newts, and frogs survived in patches; some of those patches likely held dinosaurs for a while. Re-conquering the planet from a crashed, inbred, specialized remnant is another matter. T. rex's excellence in the world it evolved for is no help once that world is gone.

Would humans have evolved if the rock had missed? Hone's guess is probably not. He does not buy Simon Conway Morris's style of inevitability, nor a butterfly-effect story in which one mammal under one T. rex foot erases us. For something like 120 or 130 million years of the Jurassic and Cretaceous, almost anywhere on Earth, an animal larger than a raccoon was probably a dinosaur — as dominant as mammals are now, or more. Primates may already have been present; molecular clocks hint at it, though osteological evidence is missing, and Purgatorius is the tiny, rodent-like form Fridman pulls up. Recent analyses, Hone says, do not show dinosaurs already in terminal decline. After they did go, giant predatory birds such as the phorusrhacids competed with mammals, so even the takeover was not guaranteed. He is impatient with the Victorian ladder that makes reptiles stupid by definition. Dale Russell's famous dinosauroid, a humanoid Troodon built for television and still copied in museums, always comes back because Troodon has a high encephalization quotient. Hone asks why intelligence would shed the tail, grow plantigrade human feet, and recast itself as a person. Crows and parrots already perform at ape-like levels on some tasks, on a different body plan. Island faunas, which produce the weirdest living animals, barely fossilize. The strangest dinosaurs are probably the ones we will never see.

What still fries Hone's brain is not a secret mechanism. It is the numbers. He talks in 70 million and 104 million the way other people talk in weeks, and he cannot conceive 66 million years any better than his students can. A friend corrected the usual dice metaphor: evolution is not rolling until you get all sixes; it is rolling and keeping the sixes. Even rare species number in the thousands or hundreds of thousands, persist for hundreds of thousands of years, and, unlike whales, apes, and elephants, often throw dozens or thousands of offspring. Ocean sunfish, Mola mola, can release on the order of 300 million eggs in a spawning. With those totals, a useful variant is not a miracle. What people miss, he says, is compromise. Darwin's phrase is descent with modification: edits to an existing plan, which is why tetrapod limbs keep one bone, then two, then a handful of blocks, then digits, and why tentacles do not sprout from nowhere. Snakes and legless lizards lose limbs; they do not invent a new drawing. Lion manes darken until heat kills. Human retinas wire backwards and leave a blind spot; squid eyes, evolved separately, do not. Tiktaalik did not sprout land life as a brainstorm. It already had a swim bladder exchanging a little oxygen, a structure built for something else, and function shifted. Traits are rarely single-purpose. Elephant tusks now dig, fight, and strip bark; the first nubs were probably for display, because they were useless as shovels.

Fridman asks the question Hone still meets on the radio: are birds dinosaurs? The answer is just yes, in the same way humans are apes and mammals. Trace living birds back, then a few million years further, and the nearest relatives sit close to a small Velociraptor. There are on the order of 10,500 to 11,000 living dinosaur species. Birds did not rise from the ashes of the extinction; they lived beside other dinosaurs for something like a hundred million years, took a kicking at the boundary, and a handful of lineages crossed. Feathers are older still. Middle Jurassic tyrannosaurs at about 165 million years — a hundred million years before the impact — already have simple, chick-like feathers. Archaeopteryx in the Late Jurassic has modern vaned feathers; Chinese Early Cretaceous beds are full of them. Hone lists eight or nine major groups as fully feathered as far as we can tell, including tyrannosaurs, dromaeosaurs, troodontids, and ornithomimosaurs. Insulation for warm-blooded metabolism and display that scales cannot match — molt, recolor, erect, differ between juvenile and adult — arrive together. As soon as you have feathers for warmth, someone can turn them red. As soon as you have them for show, you stay warmer. Biology, he keeps saying, does not assign one job per part.

Beside him the whole time is a cast he barely mentions: a juvenile *Protoceratops andrewsi* from Mongolia, pig- or sheep-sized as an adult, the one skull he can afford and the species he thinks his colleagues underrate. More than a hundred good skeletons, perhaps 70 or 80 in accessible museums, run from embryos to large adults, almost all from one Mongolian window only about 100,000 years wide. That is a population. You can ask when the display structures take off, whether the sexes differ, how juveniles die. He and students once scraped together about 110 gharials — long-snouted, highly endangered crocodilians — from museums worldwide and found that except for the largest males, females and smaller males overlap. Dinosaurs grow and stay sexually mature across a long size range, so a big female looks like a small male. Papers already disagree on whether Protoceratops is dimorphic. The trait could be strong and still invisible until someone IDs laying females by medullary bone, perhaps by drilling a mass-death assemblage from the breeding season. New species in Mongolia keep making headlines. Hone wants more Protoceratops. Another hundred of them, he says, would tell us more about how dinosaurs worked than fifty new names.

Does a career in deep time make him sit back more often and say this is incredible? He does, but perhaps not more than he would as a zoologist who happened onto dead animals. He is as ready to stare at lungfish, flying frogs, caterpillars, velvet worms, and hagfish. In Uganda he watched marabou storks circle and then remembered their wingspan is about a fifth of the giant pterosaurs he studies. Size-obsession is understandable: blue whales out-awe humpbacks, and T. rex, Triceratops, Quetzalcoatlus, and Patagotitan hold the eye. What is harder to hold is duration. Elephants have come and gone; whales reached extremes recently and nearly vanished in a century. Titanic dinosaurs lasted at least a hundred million years, and after the Triassic the planet was theirs for something like 140 million years, almost everywhere. That dominance, not any one skull you could walk through, is the fact Hone still cannot scale to a human week.