dundant use of bone to place the tallest plates over the strongest
segments of the back.
American paleontologists of the last four generations have
puzzled over the apparent incomprehensibility of the arrangement
of the stegosaur's plates. Several prominent museum scientists even
concluded there simply wasn't any mechanical function at all for
the plates—they were purely ornamental devices to make the
stegosaur look more intimidating to enemies and sexier to poten-
tial mates. A sexy look and an intimidating profile are worthy evo-
lutionary results. But in fact if the evolution of the stegosaur's plates
is carefully considered, it becomes possible to see how they could
have functioned as a very effective addition to defensive arma-
ment.
Armor plate was a long-standing characteristic in the Dino-
sauria. Crocodiles sported armor on their neck, torso, and tail when
they first appeared during Triassic times, 220 million years ago.
And all living croc species retain a flexible body shield of horn-
covered bony plates joined together by sheets of ligaments within
the deep skin layer. No croc possesses stegosaur-style triangles,
but most crocodilians do display big oval plates of armor on the
back of the neck, and these usually have raised, pointed ridges.
Similar arrangements of armor protected other dinosaur relatives
(the Thecodontia) from the Triassic Period.
The process whereby oval, keeled plates evolved into the up-
right, thin triangles of stegosaurs was fairly simple: by reducing the
bony bases and enlarging the ridges, the oval plates quickly be-
came thin stegosaur triangles. However, the key question about
stegosaur armor is why would the animals evolve such weak nar-
row bases for their plates? A broad base firmly embedded in tough
skin would have prevented the armor from bending when a pred-
ator struck. But stegosaur plates were so tall, and their bases so
narrow, it was most unlikely they could remain stiff or upright.
Maybe the stegosaur's approach to armor design was dynamic.
Muscles in the skin might have moved the plates so they could
232 | DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
point upward or downward, depending on the point of origin of
the attack. When erect, the points would have warded off attacks
from above; when flexed downward, they could have provided the
stegosaur with a very effective flank defense. Held horizontally,
the tallest plates would have stuck out sideways three to four feet,
so that no predator could approach near the stegosaur's hide. Since
the muscles and flanks of the hind limb were especially vulnerable
zones, the plates over the rump had to be the largest.
A flexible armor defense would at last explain all the most
peculiar features of how stegosaurs were designed. The pitted tex-
ture of the plates' basal surface would have provided purchase for
some sort of tissue—ligament or muscle—to embrace the plates at
the base for about half a foot. Thin bases embedded in the skin
would have been necessary to permit the hingelike movement of
the plates. The thin, triangular shape conveyed the greatest strength
to the pointed tip with the least concomitant amount of weight.
Muscles in the skin have evolved several times in different
vertebrate groups, and might easily have evolved in stegosaurs. Skin
muscles have evolved in mammals to move hair as when a horse
twitches the skin on its back to flip off a fly. (Human evolution
took a U-turn here. We had skin muscles once but lost most of
them when we evolved our naked skin, and now possess few ex-
cept on our faces.) Most living reptiles bunch skin muscles thickly
around the throat, and they are what slips the neck frill forward
on the Australian frilled lizard, for example. The skin muscles of
birds are used to control the orientation of their feathers. If skin
muscle could evolve to flip frills in lizards and feathers in birds,
stegosaurs might have evolved them to flip their armor plates.
Stegosaurus must have been a grand performer under attack—
a five-ton ballet dancer with an armor-plated tutu of flipping bony
triangles and a swinging war club. Browsing peacefully on the tops
of bushes, perched upon its hind legs and tail, its keen eye quickly
catches the movement of two huge Allosaurus, hunting au pair, along
the floodplain. The pair of allosaurs stride quickly to the attack,
one from either side. The Stegosaurus alights defensively on all fours.
The first Allosaurus darts quickly for the stegosaur's neck. In-
stantly the intended victim pivots and lowers its armored trian-
gles. The allosaur suffers a cut across its snout for its efforts. The
second allosaur lunges at the other flank, but the stegosaur's tail
MESOZOIC ARMS RACE I 233
slashes out to meet the charge. With an audible "whoosh," the four-
spiked tail barely misses the allosaur. That's too much for the would-
be predators. They back off from the stegosaur still capably bran-
dishing its weapons. And off they go to find easier pickings.
Sudden extinction interrupted the evolution of dinosaur ar-
mor at the end of the Jurassic when the true stegosaurs died out
totally or at least became very rare. But as the Cretaceous Period
dawned, new dinosaur dreadnoughts appeared: the nodosaurs.
Nodosaurs shared some characteristics with the stegosaurs—their
hips were high and their deltoid muscles were strongly developed
for sideways maneuver. But on the whole, the nodosaur's ap-
proach to defense was much more massive. A complete flexible
Four tons of
charging armor—
the nodosaurian
Edmontonia
234 | DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
pavement of small bony plates armored the entire upper surface
of nodosaurs—head, neck, torso, and tail. None of these plates were
anything like the tail triangles of the stegosaurs, but some nodo-
saur species had broad-based plates topped by tall conical spines.
Nodosaur armor plating was therefore much more continuous than
that of stegosaurs, and their hip bones had expanded into an im-
mense solid roof protecting the upper surface of the abdomen.
Some paleontologists believe nodosaurs defended themselves
passively. According to this view, these massively armored mon-
sters employed their carpet of plates much like a mobile bomb
shelter. Under attack, they would merely hunker down on the
Cretaceous forest floor, legs folded under their body, to wait out
the tyrannosaur's attack. But there is ample reason to believe no-
dosaurs could become actively dangerous antagonists and turn the
tables on their attacker. What made some nodosaurs dangerous was
their sharp shoulder spikes. Mounted on broad bony bases
embedded in the skin of the shoulder stood long, sharply pointed
horn cores that curved forward. In life, an outer covering of horn
made these shoulder spikes as long and deadly as the tail spikes
of stegosaurs
had been. Like stegosaurs, nodosaurs had very strong
elbow muscles (the triceps muscle group) perfect for quick, for-
ward lunges. Altogether, the forward end of a nodosaur resem-
bled the gigantic, short-legged warhorse of medieval times, coated
with armor, and ready to charge with wickedly sharp lances jutting
forward from either side of its head.
Wrestlers and short-legged fullbacks know the advantage of
short, strong legs—a long-limbed wide receiver can run faster, but
the shorter legs provide a greater initial acceleration. The nodo-
saur's enemies were the very long-limbed tyrannosaurs, fast enough
to catch any nodosaur very quickly. But once the combatants were
close, the advantage of speed disappeared. A tyrannosaur could
stretch downward to snap at the nodosaur's tail or back, but the
predator would only succeed in breaking its teeth against the im-
pregnable carapace. The tyrannosaur's only hope would have been
to flip the nodosaur in order to attack its unprotected belly. But
nodosaurs were very wide across the hips and had a low center of
gravity. A fully grown nodosaur would have been as easy to flip as
a modern wide-track station wagon. And then the tyrannosaurs
would have had to face counterattack. The nodosaur could have
MESOZOIC ARMS RACE I 235
Weakest teeth in the Dinosauria—the nodosaurs. Edmontonia was typical of
the strangely constructed nodosaurids. Thick armor plates covered the snout
and forehead, and a sharp-edged beak was built around the muzzle, but the
teeth were absurdly tiny. Did this dental decrepitude condemn Edmontonia to
a diet of soft-water plants?
kept pivoting to face its long-limbed attacker, awaiting its op-
portunity. If the tyrannosaur allowed any opening at all, the no-
dosaur's powerful elbows and knees could instantly drive its
armor-plated body forward. And its murderous shoulder spikes
might catch the predator's calf or leg, tripping the tyrannosaur or
ripping a nasty wound. Before such a lethal charge, any tyranno-
saur might have beaten a well-advised retreat.
Late Cretaceous days were the high point of armor develop-
ment for the Dinosauria. Nodosaurs of several species stalked the
meadows of the Cretaceous deltas of North America. And they
were joined by a new family of dreadnoughts, the ankylosaurids.
At first glance the ankylosaurids seem less dangerous than the no-
dosaurs: they were shorter at the hips and weaker in the shoulder,
couldn't pivot as quickly, and lacked the lethal shoulder spikes. As
compensation, ankylosaurs had better head protection than no-
dosaurs—overhanging plates protected their eyes and cheeks.
But the tail functioned as the cornerstone of the ankylosaur's
defensive tactics. Like the nodosaur's, the root of the ankylosaurs'
tail was powerful and supple. Unlike nodosaurs, the last half of
the ankylosaurs' tail was stiffened by a series of bony tendons that
converted this end into a prolonged handle for a bone-crushing
war club. At the very end of the tail, three large masses of bone
fused together to create a sort of monster cloverleaf-shaped club
head. By contracting its tail muscles, the animal could quickly swing
the stiff handle from side to side and powerfully flail its three-leafed
club head.
The ankylosaurs' war club was less precise but quicker than
the stegosaurs'. A stegosaur's tail joints were supple right down to
the very tip of its tail, so the animal could control the movement
of the spiked club carefully. Such care was necessary for effective
defense because the sharp spikes required accurate aim. Ankylo-
saurids had less finesse, but the massiveness of their club guaran-
teed damage no matter how the blow landed.
In nodosaurs, ankylosaurids, and stegosaurs, the dangerous
weapons were carried on shoulders or tail. But one group of ar-
mored dinosaurs, the boneheads, used their skulls as their princi-
pal offensive device. "Greatest Dinosaurian Bonehead!"—touted
the label beneath the skull in the New York museum. In the glass
case stood Pachycephalosaurus ("heavy-headed lizard"), a bonehead
MESOZOIC ARMS RACE I 237
indeed, with a two-foot-long skull topped by a dome of solid bone
eight inches thick covering the forehead and crown.
Bone-headed dinosaurs at first provoked the same combina-
tion of awe and ridicule that originally greeted the stegosaurs. Here
was yet another case of outlandish dinosaur construction without
any conceivable explanation from body mechanics. Back in the early
1900s, the theory of racial senescence would have served as a re-
spectable hypothesis to explain strange dinosaur adaptations such
as bonehead skulls. Flamboyant and senseless crests, plates, and
spines were supposedly signs that all Dinosauria had gone senile
in their evolutionary old age. Like old ladies wearing out-of-fash-
ion headgear and mismatched gowns and coats, the dinosaurs had
allegedly lost their adaptive vigor and had become incapable of
evolving anything but nonfunctional ornaments and maladaptive
excrescences. As a theory, racial senescence was bankrupt by 1920,
but it still pops up here and there in bad books about dinosaurs.
The bodies of boneheads were nothing unusual: a pair of long
hind legs, long, stiff tail, short arms, a barrel-shaped body mass to
accommodate masses of vegetation. This general configuration
nosaurs. And even bonehead skulls weren't noticeably deviant in
the snout, jaws, and teeth. The strangeness of boneheads was con-
centrated entirely in the domelike swelling over the top of the
braincase. Some species had only a slightly thickened skull roof of
otherwise normal construction. But the fully developed bone-
heads, like Pachycephalosaurus, had giant domes that suggested great
intellect. As the original discoverers of bonehead dinosaurs were
quick to point out, the supposedly brainy appearance was a sham—
the brain itself was tiny and occupied only a small volume deep
inside the bony dome. Most of that dome was in fact filled with
bone cells arranged in a radiating pattern like the fibers in a cross
section of grapefruit. Bone cells usually grow in the direction of
greatest stress, so the bonehead's pattern of growth is a clue to the
head's function. The radiating pattern strongly suggests that the
dome was subjected to enormous outside pressures. But of what
sort?
Dome-headed dinosaurs can probably best be understood as
wearing NFL-style football helmets over their minuscule brain-
case. Modern football helmets weren't designed for merely pas-
238 | DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
Stegoceras—two individuals ramming
sive protection; they were built so the wearer could ram his head
into the unfortunate player opposite him. Old-fashioned leather
helmets weren't as good for the head-first block, but coaches soon
discovered that the human head could serve as a weapon, so
the
helmets were redesigned. A domed, impact-resistant helmet was
invented to provide the head-ramming linebacker with the ability
to strike blows without damaging his own skull. Dome-headed di-
nosaurs had evolved the physical equivalent of a ramming helmet
millennia before, a configuration that strongly suggests a head-first
mode of attack.
Peter Galton, a leading expert on beaked dinosaurs (he's the
man who studied the cheeks and digestive tracts in duckbills), first
worked out the head-butting theory in 1971 and won nearly
worldwide acceptance of his basic idea. Galton pointed out that
dome-headed dinosaurs had exceptionally strong muscles holding
the head at a right angle to the neck, so that the dome would face
forward when the beast lowered its head and charged. The animal
possessed all the qualities for an optimal butting attack—a bull neck,
MESOZOIC ARMS RACE I 239
a low head position in the charge, and a thick skull covering a small
brain.
Galton believed sex was the ultimate motivation behind the
head-butting behavior. He argued that male domeheads would have
banged their crania against one another in ritualized combat, much
like bighorn sheep. On this point, I would disagree a bit. A big-
horn sheep's horns are wide and flat, so when two males clash, their
horns meet across a wide surface. The resulting collision is a true
test of strength, because the full force of the sheeps' bodies is de-
livered. But the rounded shape of bonehead dinosaur domes made
a precise head-to-head blow nearly impossible. If two boneheads
did clash, their heads would probably have struck only glancing
blows. Domed heads, therefore, like football helmets, were prob-
ably for butting an adversary in the body, not in the head.
Polish expeditions to Mongolia in the 1960s found a marvel-
ous bunch of bonehead dinosaurs. The Poles wondered whether
the head-butting equipment really was for sex-related contests—
the head seemed too dangerous a weapon for such encounters. The
Polish hypothesis argued that the bonehead was essentially an an-
tipredator weapon. It is difficult to judge between the alternatives.
Did the boneheads ram one another, or did they ram meat-eating
dinosaurs that threatened attack? Probably both hypotheses are