ing. This sort of dominance of the upper teeth was a characteristic
of the dinosaurs; Allosaurus, Ceratosaurus, and Tyrannosaurus all had
much larger uppers than lowers.
The biological engineering behind Coelophysis's bite can be
worked out from well-preserved jaw joints and the muscle-attach-
ment sites they reveal. The shape of Coelophysis's teeth indicates
the upper row of teeth had to move rearward relative to the lower
so the bigger crowns of the upper teeth could be effectively ex-
ploited. Since it consisted of a pair of grooves that allowed two
262 I DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
knobs on the skull to slide, Coelophysis's jaw joint was indeed ar-
ranged so the skull could shift fore and aft in relation to the lower
jaw, and the biggest jaw muscle pulled the snout backward. The
strong neck would also assist in delivering a killing blow because
Coelophysis could rake its teeth backward through its prey by re-
tracting its neck muscles. Such contractions of the neck were am-
plified by the back of the skull, which was strengthened and
enlarged to support larger and more powerful neck muscles.
From such evidence, the lethal interplay of predator and prey,
anchisaur versus theropod can be fairly clearly imagined. Coelo-
physis stalks the Triassic floodplain, head held high, its large, bird-
like eyes scanning the landscape for the slightest movement. A
rustle in the conifer bushes betrays an anchisaur, a small, half-grown
specimen some four feet long. Coelophysis strides to the attack. The
anchisaur, its back to a dense stand of undergrowth, rises on its
hind legs, brandishing its foreclaws. The combat is joined. Coelo-
physis dances, darting in and out in feinted strikes, weaving to avoid
the dangerous counterstrokes from the anchisaur claws. An open-
ing appears—perhaps a momentary error on the anchisaur's part.
And Coelophysis lunges, its tooth-studded jaws raking some ex-
posed part of its victim.
In a split second the entire series of the predator's jaw and
neck muscles fire off in a spasmic, contractile sequence originating
in instinctive action unguided by conscious thought. The anchi-
saur struggles to free itself but its efforts serve only to make the
wound longer and more ragged.
The predator's dance continues, punctuated by more feints and
quick raking strikes. No one bite is fatal. There is no quick coup
de grace like a lion's. But Coelophysis's prey succumbs after a short
time, weakened by trauma and loss of blood. Finally, the anchi-
saur sinks to the ground, unable to right itself, and Coelophysis
finishes with a series of slashing bites to the neck just behind
the head.
Coelophysis was not the only practitioner of this style of hunt-
ing at the end of the Triassic and beginning of the Jurassic. Sam-
uel Welles of the University of California hunted in the red beds
on a Navaho Indian Reservation and found several nearly com-
plete skeletons of big predators, between fifteen and twenty-five
feet long. Today, these are the earliest complete skeletons of large
predatory dinosaurs known. Welles's animal, Dilophosaurus, "two-
DEFENSE WITHOUT ARMOR | 263
crested-lizard," exhibited a striking similarity to Coelophysis in its
very long tail and elegant hind limbs. But the two-crested dino-
saurs were proportioned for killing much larger prey, with their
shorter, more massive necks and skulls and very much larger up-
per teeth relative to the skull's length. These dinosaurs were strong
enough to attack any of the Early Jurassic herbivores, even the
largest anchisaurs.
As the long Jurassic Period passed through its middle and late
epochs, the dinosaur arms race produced more heavily armored
herbivores—the stegosaurs—and the immense brontosaurs with
enough strength in their legs and feet to simply crush most pred-
ators. Predator strength increased too; the Late Jurassic Ceratosau-
rus was thirty feet long, and Allosaurus forty-five. Ceratosaurus and
Allosaurus were both discovered by Professor Marsh in the late
1870s. And for a long time only one ceratosaur's skull and only
two or three complete allosaur skulls were known. Then, in the
1940s, a spectacular predator trap, containing ceratosaurs and al-
seventy Allosaurus specimens at all stages of growth—young, adult,
aged—have been quarried from this small area of mudstone. Jim
Madsen, state paleontologist of Utah, directs the work at the quarry,
and his practical experience with hundreds of predator bones en-
dows him with unequaled expertise on the subject of predator
anatomy.
I have spent several unforgettable weeks in Salt Lake City
studying Jim Madsen's laboratory full of allosaur and ceratosaur
bones. In this astounding treasure house every detail of their bio-
mechanics stands revealed. A most unexpected characteristic of the
skulls is how easily they fall apart. A fully adult Ceratosaurus's skull,
nearly three feet long in life, was not one tight mass of bones and
teeth; it consisted of a loose kit of thin bony struts, flexible bony
sheets regularly perforated by holes, ball-in-socket joints, and sliding
articulations, the whole bound together with ligaments. After death,
the ligaments of course soon rotted and the skull fell apart, scat-
tering its pieces across the mud. Today's largest predatory mam-
mals—polar bears and lions—possess a strong, unified cranial
structure that remains solid long after death. Bioengineers who
study skulls must consequently refashion their thinking when they
seek to reconstruct the mechanics of the loose rod-and-sheet con-
struction found in Ceratosaurus and Allosaurus.
264 | DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
Skull of Ceratosaurus
There was a strong central core in the heads of the predatory
dinosaurs: their thick-walled braincase. The term "braincase" is a
misnomer in dinosaur anatomy, because in fact the brain of larger
species was minute compared to the surrounding mass of bone.
The primary function of the dinosaur's "braincase" was to provide
attachment sites for the neck muscles and to serve as the founda-
tion point for all the thinner, more flexible components of the snout,
palate, and roof of the skull.
The biggest surprise found in Madsen's ceratosaur skull was
the tooth-bearing bones of the snout. Instead of being firmly at-
tached to the braincase, the tooth-bearing bones were only loosely
bound to the top of the snout and the roof of the mouth. Such
looseness is repeated all through this skull. The tall strut of bone
(called the quadrate) which connected the lower jaw to the brain-
case shared a hinge joint with the top rear corner of the skull. When
this strut swung outward, it splayed out the jaw to the sides. Even
the lower jaw was loosely constructed of two sections. The front
section carried the teeth, the rea
r housed the muscles and joint of
the jaw. The front and rear complexes met along a quite loose lig-
amentous junction. At the dinosaur's chin, the right and left lower
jaws met at yet another very weak joint held together by liga-
ments.
So much looseness was baffling to biologists who knew only
the mechanics of our own Class Mammalia. If we humans had as
DEFENSE WITHOUT ARMOR I 265
How to swallow something
larger than your head—
dinosaur-style. Face-front
view of Ceratosaurus. All the
bones of the skull's side
were loosely hinged to the
skull top, so the head
expanded sideways when
the beast swallowed an
extra-large meat chunk. And
a hinge in each lower jaw
opened outward, just like a
boa constrictor.
loose a skull as the ceratosaur, every time we bit down, our
cheekbones would flex inward, the roof of our mouth would con-
tract, and we would feel the rear of our skull swing toward the
base of our neck. Anyone who has kept snakes as pets wouldn't
be puzzled by ceratosaur heads. The heads of snakes are generally
similar in design to those of the dinosaurs—snakes have a central,
tightly knit braincase, which acts as the core for the loosely at-
tached jaws, snout, cheek bones, and palate. Snakes also possess
backwardly curved teeth, another similarity. When a snake starts
266 | DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
to swallow large prey, the jaw muscles pull these teeth into the
prey's body and all the loose joints swing apart so that the snake's
gullet can accommodate a very large body. The Ceratosaurus must
have functioned in very much the same way. When a ceratosaur
swallowed a large chunk of meat, its capacity would have in-
creased as each loose joint flexed and bowed outward.
The largest prey commonly available to Allosaurus and Cera-
tosaurus were the stegosaurs and the brontosaurs. Stegosaurs of
course wielded their spike-and-plate armor, but at first sight bron-
tosaurs appeared poorly armed. Most brontosaurs did have short,
inwardly curved claws on their fore and hind feet, so the paws were
potential weapons. But a more potent defensive weapon was lo-
cated at the rear end of the whip-tailed genera like Brontosaurus
and Diplodocus. The final ten feet of the tails of these dinosaurs
featured slender bony rods in the core of the tail. When these huge
dinosaurs swung their hugely muscled tails, the whiplash effect could
inflict crippling wounds on an unwary predator.
The ultimate phases of the arms race between predator ar-
mament and antipredator adaptations were played out during the
Cretaceous. Allosaurus, itself a Late Jurassic type, displayed the
beginning characteristics of Cretaceous-style hunters, while Cera-
tosaurus represented the older predatory design, little changed from
Early Jurassic days. The Allosaurus's skull was more thickly boned
than that of Ceratosaurus, and its jaws were deeper, providing for
larger jaw muscles and a larger, stronger area for neck muscles.
Not only was Allosaurus's bite stronger, it was also faster on its
feet. The allosaur's hind legs were longer and more compact than
those of Ceratosaurus. And from an Allosaurus-type ancestor de-
veloped the last major group of big predators: the most strongly
jawed, and fastest runners of all, the Tyrannosauridae of the Cre-
taceous.
In a glass case on the fourth floor of the American Museum
of Natural History in New York resides the single most famous
dinosaur head in the world—the Tyrannosaurus rex from Hell
Creek, Montana. All the biomechanical trends started in Allosau-
rus culminated here. Primitive theropods like Ceratosaurus had teeth
that were big but delicate and thin in section. Tyrannosaurus's teeth
were gigantic and very thick, capable of resisting exceptional forces
when biting. Whereas the ceratosaur's head was a loose strut-and-
DEFENSE WITHOUT ARMOR I 267
ligament construct, Tyrannosaurus's skull was one unified whole,
very solidly constructed, with no moving parts except at the joint
of the jaw. The compartments in the tyrannosaur's skull and in the
lower jaw that housed the muscles were enlarged more than in any
other predator. Its neck too represented an apogee of power. Ty-
rannosaurus had surrendered nearly all the primitive expansion
points in the skull. But it compensated in the lower jaw, where
the hinge between the front and back sections was much better
developed than in the older predators. When Tyrannosaurus bolted
down huge pieces of meat, the deep lower jaw flexed easily from
side to side to widen its gullet.
Tyrannosaurus and its close kin Albertosaurus (named for the
Diplodocus defends itself with
tail swipes at two allosaurs.
268 I DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
Canadian province) confronted the most heavily armored and armed
adversaries—the tanklike nodosaurs and ankylosaurs and the dan-
gerous horned dinosaurs already described. Part of any predator's
advantage is the opportunity to make feints and lunging attacks.
But such tactics require a good judgment of space and distance,
DEFENSE WITHOUT ARMOR I 269
Tyrannosaurus rex,
five-ton predator
of seventy million years ago
and early predatory dinosaurs possessed very little depth percep-
tion, because their eyes faced directly sideways. The tyrannosaur's
snout was sharply pinched to clear its field of vision. And its eyes
faced forward to provide some overlap between visual fields from
the right and left eyes. That would have permitted stereoscopic
vision. Moreover, evolution had made additional improvements for
attacking dangerous prey in the tyrannosaur's limbs. Its hind leg
was much longer and more compact even than Allosaurus's. And
its torso was shortened to benefit balance and speed. Despite its
great size—up to five tons— Tyrannosaurus was surprisingly slender-
limbed, graceful, and fast.
All these evolutionary increases in the bulk of its jaw muscles
and the strength of its limbs seem to demand that something be
270 | DEFENSE. LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
Leg proportions in a tyrannosaur,
Albertosaurus
eliminated from the tyrannosaur's design. Forelimbs had to go.
Ceratosaurus had had short but well-muscled forelimbs. Allosaurus
had had shoulders of reduced bulk but still had had a strong hand
and a fearsome claw on its thumb. Tyrannosaurs reduced their
forelimb to such an extreme that it appeared useless, or nearly so.
A thirty-foot Albertosaurus's arm was shorter than a man's, and most
of the muscle-attachment processes were subdued. So the hand was
not only short, it was weak. Strange as it may sound, any average
adult human could have won an arm-wrestling contest with a five-
ton Tyrannosaurus.
Additional weight
was saved in the tyrannosaur's hind foot.
Very early predatory dinosaurs had had strong claws on the three
main toes. But the tyrannosaurs reduced both the size of the claws
and the bulk of the tendons and muscles supporting them. Their
feet were thus adapted for running and dodging, avoiding coun-
terattacks from the spikes, tail clubs, and horns of their prey. Strong
hind claws might have been useful weapons but their weight would
have detracted from speed and nimbleness. Tyrannosaurus surren-
dered the attack function of both the hind and forefoot in favor
of a concentrated mass of muscles and power in the neck and head.
A final mystery looms large in the story of predator and prey.
At present I can offer no solution for this and neither can anyone
else. In most places, the most common, large plant-eaters of Late
Cretaceous days weren't the heavily armed horned dinosaurs or
the armor-clad ankylosaurs. Most common were the naked-skin
duckbills, which lacked any sort of obvious defensive weapons.
Duckbills had no whiplike tails, long claws, or any type of spike
or plate. And their limbs were shorter and designed for lower top
speeds than were those of their gracefully long-legged hunters. How
ever did duckbills escape their enemies? To date, no one knows.
But I am convinced some young paleontologist, perhaps someone
reading this book, will one day solve this enormous riddle.
272 | DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
13
DINOSAURS TAKE TO
THE AIR
Seventy million years ago a dragon of the air stretched its mem-
branous wings over the Texas delta. Forty feet from wingtip
to wingtip, this aerial leviathan possessed a wingspan greater than
some twin-engine airliners and was three times wider than the
greatest living bird, the Andean condor. The fossil annals in the
Texas rocks yield an image as marvelous as any fabrication of the
human imagination. Petrified wing bones, vertebrae, and jaws make
it possible for us to envision the largest flying creature produced
by evolution.
Flying dragons entered the sphere of human knowledge not
as giants, but as tiny winged skeletons from the fine-grained lime-
stone of Bavaria. There, quarrymen hewed out slabs carefully, be-