The Nanotyrannus fossil first came to the attention of paleontologists in 1942, when it was unearthed by an expedition of Cleveland Museum researchers digging in the plains of Montana. At first the researchers didn’t know precisely what they had found, but their initial examination seemed to indicate that the find was a rather small member of the carnivorous dinosaur genus Albertosaurus. (Actually, at the time the genus was known as Gorgosaurus, but it has since been renamed, after it was found to be identical to a dinosaur already unearthed in Alberta, Canada.)

Albertosaurus was a primitive relative of Tyrannosaurus rex and it was distinguished from its more famous cousin mostly by size. T. rex measured up to 50 feet from snout to tail and could tip the scales at as much as 14,000 pounds. Albertosaurus was a more compact 25 feet and 4,000 pounds; the skull dug up by the Cleveland group would have belonged to a runt, one measuring 15 to 20 feet and weighing 1,000 to 1,500 pounds.

For all its albertosaurian features, however, the Cleveland skull always bothered me. From the first time I saw it, I was troubled by its small size, as well as by its long, narrow snout and its unusual, forward- looking eyes. It was like going to the zoo and seeing a jaguar in a cage marked horse. In 1987 I got a chance to examine the puzzling Cleveland skull and concluded that it did not come from an albertosaur at all, but from a whole new species of tyrannosaur that Mike, Phil, and I dubbed Nanotyrannus, or pygmy tyrant.

Whatever its name, though, the specimen clearly seemed to belong to the dinosaur grouping known as carnosaurs (literally, flesh lizards). The carnosaurs include all the other big meat-eating dinosaurs, such as Tyrannosaurus, Ceratosaurus, and Allosaurus, among many others. These are the large predators that are so popular with schoolchildren. But paleontologists have always had a prejudice against them. Carnosaurs were thought of as big and fearsome but incomparably dumb. Slow-footed, Godzilla-like behe- moths, they represented a muscle-bound dead-end in the history of life, a Schwarzeneggerian cul-de-sac in evolution that left no living descendants.

Much more interesting were the carnosaurs’ pint-size cousins, the coelurosaurs (hollow lizards, a reference to their hollow tailbones). Weighing between 10 and 200 pounds, these little predators had sharper senses, greater agility, and much greater speed than their lumbering relatives. Their advanced design would make them better adapted to a lively mode of existence.

Perhaps the coelurosaurs’ most impressive feature, however, was their striking resemblance to today’s birds. Many of the animals’ features- -for example, their short torsos, long hind legs, reversed inner toes, and reversed pubic bones--are so unmistakably avian that paleontologists see the ancient coelurosaur line as leading almost indisputably to modern birds. Apparently, while the carnosaurs were destined for evol- utionary oblivion, the coelurosaurs were destined for the skies. The two lines, it was generally agreed, were quite distinct and had probably evolved separately from the very earliest days of dinosaurs, back in the Triassic Period, which ended 213 million years ago.

The idea of an unbridgeable gulf between large and small predatory dinosaurs was established early in this century by famed paleontologist Henry Fairfield Osborn, who helped establish the magnificent fossil collection at the American Mu- seum of Natural History in New York. In 1905 Osborn discovered the first T. rex, dug from Cretaceous sediments about 65 million years old. He interpreted the monster as the ultimate Darwinian development of the earlier Allosaurus and Ceratosaurus.

Even though the T. rex was a crowd-pleaser, Osborn thought that several smaller dinosaurs he had found were much more momentous discoveries. They went by names like Ornitholestes (bird robber), Struthiomimus (ostrich mimic), and Saurornithoides (liz- ard with a birdlike shape)--nimble, bantam-weight creatures that, as their names implied, were extremely avian in appearance.

Osborn believed that all these creatures came from a single line that also gave rise to Archaeopteryx--universally accepted as the first true bird--145 million years ago. Archaeopteryx descendants then evolved into the birds we know today, while the remaining coelurosaurs retained their birdlike traits (though they, of course, never left the ground).

This idea fell in and out of vogue until the 1970s, when it was given a boost by Yale paleontologist John Ostrom. Ostrom conducted extensive studies comparing the hands of Archaeopteryx with the hands of the coelurosaur Deinonychus and found them to be variations on a similar theme. Throughout the 1980s the coelurosaur-bird link was the one of the sexiest subspecialties in the field of dinosaurology--and the one most likely to get you job offers and grants. (One of our colleagues, who had done superb work on fossil lizards, freely admitted that he converted to birdlike dinosaurs because, he said, Dead lizards won’t get me a salary.)

However, the coelurosaur reputation for alacrity didn’t rub off on their giant carnosaur kin. Allosaurs and tyrannosaurs still had the stigma of being too big and too slow, with brains that were too underpowered. Mike, Andy, and I, though, were not so sure that carnosaurs should be excluded from the evolutionary fast lane. We agreed with the general principle that predatory dinosaurs and birds were close evolutionary kin, but we thought that segregating the big carnivores from the small carnivores simply because of body bulk and inferred agility was arbitrary at best, dead wrong at worst.

In 1988 we began trying to find out just how valid the distinction was. One of the first and most telling signs of the relatedness of birds, carnosaurs, and coelurosaurs, we knew, would be the neurological wiring pattern inside their skulls. Nerve sites don’t change easily or often in evolution, so when they do, they provide a good indication that one major species group has actually diverged from another one. There are a dozen nerves that exit from the braincase bones of birds and some dinosaurs, and we believed that the way the exit holes were arranged would provide a lot of information about the way the species evolved. To probe the holes, I chose a simple research tool: the coat hanger.

Working with a T. rex skull from the Tyrell Museum, I straightened the hanger and bored out the rock filling the exit holes. What I was looking for was an exit site for a nerve labeled V1. In all modern- day vertebrates except birds, this nerve exits sideways through a large braincase hole, accompanied by two other nerves. Then the V1 turns forward to run all the way to the snout, where it operates as a tactile sensor. (When you tap your upper lip hard below your nose, you are stimulating the V1.) In birds this arrangement is different: the V1 exits the skull alone through an independent canal passing forward through the braincase bones. No one is certain why the V1 gets its own exit site in birds, but we were glad it does. If we detected a similar feature in dinosaurs it might help us identify the evolutionary path of the ancient beasts.

As it turned out, I found that in T. rex the V1 exit was bored precisely to avian specifications. We had expected to see it in coelurosaurs from the Cretaceous; finding it in carnosaurs, however, was a surprise. Further investigations showed that no Jurassic dinosaur had the avian exit arrangement, so this character seemed to mark one cluster of advanced Cretaceous families, some big, some small.

Over the next two years I continued my low-tech study of skulls, cranially abusing any number of large dinosaurs, while Phil conducted similar examinations of members of the Troödon family, small, slender- snouted, ostrichlike predators that include Osborn’s Saurornithoides. To examine the skulls of these creatures, Phil used a technique similar to mine but relied on fine needles instead of crude coat hangers. Meanwhile Andy took a far more sophisticated route, employing a tool never before used in the study of dino bones: the CT scan.

After borrowing an albertosaur skull from the Royal Ontario Museum, Andy obtained access to a local hospital CT scanner and passed the fossil through it. As we expected, the coat hanger and the CT scan yielded the same results: the V1 wiring pattern was indeed avian. Soon we were running other dinosaur skulls through the scanner, including the troödonts and the unusual specimen from Cleveland that we had named Nanotyrannus.

What we noticed right away was that, in addition to having an avian V1 arrangement, all the creatures were airheads. Modern birds have air chambers carved into their skull bones, and wide air ducts connect their cranial chambers to their windpipe and lungs (birds have air chambers nestled next to the intestines and liver too). Again, while such avian traits were not unexpected in advanced coelurosaurs, conventional wisdom said they shouldn’t be present in carnosaurs. However, the CT scans revealed that both the albertosaur and the nanotyrannosaur could match the coelurosaur in airheadedness. Further study of skull fragments from the T. rex revealed that even this mighty five-ton killer, which we’ve come to think of as the archetypal carnosaur monster, had pneumatic plumbing inside the skull.

Not only did this natural ductwork lead us to believe that carnosaurs, advanced coelurosaurs, and birds were more closely related than anyone had imagined, it also suggested that both big and little dinosaurs shared an unexpected feature: hot-bloodedness. One purpose of cranial air passages could be to help keep internal tissues cool. However, unless an animal’s metabolism runs hot in the first place, there’s no need for such organic air-conditioning.

As we compared cranial plumbing in skull after skull, we became convinced that the fundamental split between coelurosaur and carnosaur, believed in by Osborn a century ago and by most scientists since, had probably never happened. A little 20-pound Saurornithoides and a giant T. rex seemed to be small and large versions of one and the same advanced dinosaurian stock.

So we started coming up with a heterodox version of carnivorous dinosaur history. According to our new way of thinking, meat-eating dinosaurs evolved like this: An evolutionary improvement would suddenly appear in a species, and right away the new design would be incorporated into a wide array of descendant species, some very large, some very small. Repeated many times over 160 million years, this process was like a multiple skyrocket used on the Fourth of July--the main rocket keeps on going vertically, giving off sideways bursts of light every 100 feet.

Approximately 140 to 160 million years ago birdlike innovations appeared in the burst and expressed themselves in large and small dinosaurs alike. Suddenly a whole range of animals exhibited airheadedness, solitary V1 nerves, and other avian features. Some of those creatures, perhaps including Archaeopteryx, actually were birds, while others stayed on the ground, keeping their birdness in terrestrial mode. The descendants of these bird-dinos--Nanotyrannus, Troödon, and others--would die out at the end of the Cretaceous, while the true birds would fly on into the evolutionary future. The theory was not unlike Osborn’s, but it included one important difference: the avian innovations that he thought to be the exclusive province of the smaller, fleeter animals were, in fact, distributed in a much more egalitarian way. The idea that the little coelurosaurs had a monopoly on all these desirable features was quite likely a myth.

Of all the skulls we studied, from both the large and small flightless birds, by far the most startling was that of Nanotyrannus. The CT scans not only confirmed that this was indeed a new species of tyrannosaur, they also suggested that it was an extraordinarily high-tech one, possessing some traits seen only in coelurosaurs and some seen in no other dinosaurs at all. For example, earlier skull studies had revealed that all tyrannosaurs had a characteristic arrangement in which the vertical bone that makes up the midline bulkhead between the eyes is tilted up. But our scans showed that in the Nano-T this midline bone, known as the parasphenoid, goes straight forward like a long, sharp spearpoint; this is exactly the arrangement that we see in Saurornithoides and Troödon.

More important than its Troödon parasphenoid, however, was the Nano-T’s Troödon brain. There’s a widespread misconception (helped along by Gary Larson’s delightful Far Side cartoons) that all dinosaurs were pea- brained. It’s true that T. rex had a small brain compared with that of an elephant or a rhino of the same weight. But since the 1970s a wide array of smaller dinosaurs have been shown to have big brains, as large as those of birds of the same body bulk.

How do we know how big their brains were? Although brain tissue is the first to rot out of a carcass after death, the brain can leave indelible marks on fossil bone, revealing much about its size and shape. Big-brained dinosaurs had brain hemispheres that fit tightly into the braincase bones, so a clean, clear imprint is left on the inner bone surfaces. Small-brained species, T. rex included, had small, loose-fitting brains, covered with a thick layer of connective tissue. The interior of their cranium thus doesn’t show the brain shape clearly. Before we peered inside the head of Nano-T, no big dinosaur (over 1,000 pounds) had been found with a large, tight-fitting brain. As the Nano-T skull had its inner secrets exposed by the CT scanner, however, it was obvious that this species was the exception. The inner braincase bones showed a snug cerebral fit. This tyrannosaur had been outfitted with formidable mental equipment.

The list of Nano-T’s unexpectedly advanced features didn’t end with braininess, airheadedness, and bird nerves. Even more surprising was its snout architecture. The scan showed paper-thin sheets of bone running along the inner snout wall, curled around like bony cannoli (the Italian pastry). No one had seen such clear evidence of these structures, known technically as turbinals, in dinosaurs, but they are standard in some modern mammals. Dogs and hyenas, for example, owe their acute sense of smell to turbinals covered by thin layers of sensory tissue; the curl of the bony sheets increases the surface area of tissue that can be packed into the snout. To accommodate such a razor-sharp smell apparatus, the modern mammals have huge olfactory bulbs in their brains. Our CT scans revealed that the Nanotyrannus was similarly well-equipped: the braincase area was spacious enough to contain an olfactory bulb ten times bigger than that of the average modern bird, relative to head size.

The Nano-T was equipped with other superlative sensory hardware as well. Bones in the inner ear appear to have been surrounded by air chambers that would have increased sensitivity to low-frequency sounds made by rustling prey. Grooves on the cheekbones showed that the outer ear canals wrapped around the head so that both ears pointed forward. Like an old-fashioned hearing trumpet, these canals carried in sound from the front, providing the animal with stereo hearing that would have made it much better able to pinpoint the distance and direction of prey.

The Nano-T’s vision may have been similarly sharp, as indicated by its huge owl-like eye sockets and the prominent optic lobes of its brain. Since the brain wiring was very birdlike, it’s also reasonable to assume that Nanotyrannus had the full range of color vision possessed by hawks, eagles, and other modern birds.

And once the animal’s acute sense of smell, hearing, and sight helped it apprehend prey, its unusual jaw architecture would have let it dispatch what it caught very efficiently. Because of the extreme narrowness of the Nano-T’s muzzle, anything the animal bit down on would likely droop out over either side of its mouth. While this meant that less food would be contained in any one mouthful, it also meant that both of the animal’s parallel rows of teeth would get hold of the prey, greatly increasing gripping power. What’s more, the CT scans revealed that the Nano-T’s tooth crowns were nearly all the same height and were packed together tightly with few spaces in between. This would have provided an efficient, pinking shears- like bite, far more effective than what the T. rex could achieve with its irregular, gap-filled smile.

We may never know precisely which meat-eating dinosaur was the closest to modern birds, though more CT scans from T. rex and other large predators will help. For now, however, Nanotyrannus--with one foot in the carnosaur camp, one foot in the coelurosaur camp, and an eye cocked toward the avian future--is almost all the proof we need to claim that the old distinctions between big dinos, small dinos, and their feathered descendants must be discarded as obsolete.

A living Nano-T must have been an extremely alert and deadly predator. Scanning the Cretaceous meadows and forests for the faintest movement, sound, or scent of prey, the Nano-T represented the highest technology Darwinian processes had built into a carnivore. Of course, to the unfortunate creatures that populated the Cretaceous world alongside Nanotyrannus, arcane questions of speciation, relatedness, and genetic adaptations would have been utterly beside the point. As far as they were concerned, the swift, fierce, sharp-eyed predator stalking the ancient plains was simply the roadrunner from hell.

Leave Them Bones Alone

The study of dinosaur bones has always presented paleontologists with a dilemma. In order to learn everything a fossil has to tell us, we have to look inside it. But in order to look inside, we have to cut it, smash it, or otherwise mutilate it. In almost no other area of science do researchers so often find themselves tempted to destroy the very thing they are studying.

In most cases paleontologists resist this temptation and let their fossils preserve both their integrity and, alas, their secrets. Sometimes, however, we do show less restraint. One eminent nineteenth- century researcher would accidentally drop his fossil specimens to the floor to afford him a look at their insides. Another commonly practiced technique involved grinding away the specimen under examination, layer-by- layer. As internal details were revealed, they were photographed or sketched. By the time the ex- amination was finished, of course, the fossil itself had been ground to dust.

The appearance of X-ray technology early in this century provided an alternative to such practices, albeit an imperfect one. Traditional X- ray images are two-dimensional, meaning that they show all internal structures superimposed on one another. An X-ray lets you see the appearance of internal features, but not their juxtaposition.

In the early 1970s an X-ray technique known as com- puted axial tomography (CAT or CT) was developed, which provided a nondestructive way to get a three-dimensional look at the interior of an object. A CT scan is a rotating X-ray that moves around a stationary subject, viewing it from 360 degrees. The heart of the machine is a doughnut-shaped device contain- ing a single X-ray source on one inside wall and hundreds of detectors arrayed in an arc across from it.

When an object is placed inside the ring, the X-rays pass through it, fanning out slightly, and get picked up by the detectors. As the doughnut begins to rotate, hundreds of thousands of such X-ray scans are taken. When all these readings are assembled, they provide a cross- sectional image of the object, just as if a slice had been cut from it. The object then slides farther and farther into the ring, allowing other X-ray slices to be taken until a top-to-bottom scan is complete.

Ingenious as the scanner is, it still has one significant problem: most people can’t mentally rebuild all these slices. It can thus be difficult to get an understanding of how all the two-dimensional images fit together into the three-dimensional object. To overcome this, we use computer imaging systems that can reassemble the slices into a single on- screen picture; the picture can then be rotated and viewed from all perspectives.

Another wonderful feature of the CT scan is that it is extremely sensitive to the different densities in an object, all of which it represents with different levels of X-ray attenuation. In a specimen like the Nanotyrannus skull, we have been able to see fine structures like nerve passages and semicircular canals in the ear, as well as grosser, denser bone structures.

To peer through the very densest bone, paleontologists need a CT system with lots of power. Medical CTs have limited power ranges to prevent them from cooking their mainly human patients. For the biggest fossils, science must turn to industrial scanners, many of which are powerful enough to see through several feet of metal.

Of course, even with the availability of this hardware, we still do have the problem of sweet-talking the owner of the system into letting us stick a fossil into it, and then coming up with the money that we’ll no doubt be asked to pay for the privilege. Nevertheless, if the alternative is having an accident with your fossil so you can look inside, most paleontologists are at least willing to try.