during the dreary days of the 1920s when English and American
dinosaurology was settling into a muddle-headed and lackluster
orthodoxy. Nopsca was a genuine Transylvanian aristocrat with a
fondness for dressing in Balkan native dress and for reconstruct-
ing the sex lives of dinosaurs. He was also gay, a life style quite at
odds with the macho frontier style of American bone diggers.
Nopsca was universally recognized as a creative thinker—at his
worst no less thoughtful and no less accurate than muddling or-
thodoxy, at his best head and shoulders above his contemporaries
in discerning evolutionary patterns.
The hallmark of the
Thyreophoroids. All
the armored dinosaurs
were specialized in
losing the obturator
prong, a short flange of
bone that connected
pubis and ischium. The
nonarmored dinosaurs
retained the prong.
MESOZOIC ARMS RACE | 253
Nopsca argued the case for an evolutionary unity among all
armor-clad Dinosauria. Testing his evidence in Geologica Hungar-
ica in 1928, he coined the term "Thyreophoroidea"—shield car-
riers—for all armored and spiked dinosaurs. Americans more or
less ignored his hypothesis, but since 1975 there has been a Thy-
reophoroid revival. Walter Coombs, an expert on ankylosaurs,
pointed out that boneheads, stegosaurs, and ankylosaurs all shared
a most unusual feature, armored eyelids. All three groups had stiff
plates of bone embedded in their upper eyelid to protect the eye-
ball from attacks delivered from above (only the accessory eyelid
was armored; the inner eyelid was soft skin and could close over
the eye). Such armored eyelids support the baron's theory. The
most primitive beaked dinosaurs lacked any such wide, bony eye-
lids, so all dinosaurs with them could have inherited their armored
blinkers from one common ancestor.
If all the shield carriers, domeheads, ankylosaurs, horned di-
nosaurs, and stegosaurs were related, then evidence of this pedi-
gree should be found in body architecture. And we do find clues.
All the primitive beaked dinosaurs' lower hip bones (pubis and is-
chium) were joined together by a short shelf of bone, called the
obturator prong. But all the later shield carriers lacked this telltale
shelf—evidence that perhaps one common ancestor had done away
with the obturator prong when it diverged from the primitive
beaked condition. Other clues of common pedigree can be found
in the skull. In the most primitive beaked dinosaurs, the bones of
the roof of the mouth (palate) were loosely connected. But in all
the shield carriers, the skulls were far more rigid and the palate
bones firmly connected to one another.
Altogether, the baron's hypothesis now seems a happy sug-
gestion indeed. The armor-clad "suborders" probably were evo-
lutionary cousins, descendants of one branch of dinosaurs that
embarked on the adaptive path leading toward armored resis-
tance, passive and active, against the threats of the meat-eaters. The
Early Jurassic Scelidosaurus may be close to the Thyreophoroid stem.
And so the baron's term, "Thyreophoroidea," should be resur-
rected as the appropriate label for this grand tribe of armored di-
nosaurs.
254
DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
12
DEFENSE WITHOUT ARMOR
T hroughout their entire history, dinosaurs and their prey co-
I evolved in a mutually stimulating arms race. A new defense plan
among the plant-eaters would give rise to a new mode of attack
among the meat-eaters. In the previous chapter we met the ar-
mor-clad tribes. Here we shall review the parade of unarmored
plant-eaters and the evolution of their defensive equipment. These
dinosaurs with naked hides defended their vulnerable bodies with
slashing claws and lashing tails against wave after wave of meat-
eating species.
epochs of the Triassic and Early Jurassic periods. These were the
long-necked anchisaurs, distant uncles of the brontosaurs. Anchi-
saurs displayed no body armor, but they wielded huge curved claws
on their powerfully muscled thumbs and long pointed claws on their
stout hind feet. These plant-eaters therefore had defensive claws
both front and rear, a combination unusual today. So wrestling with
an anchisaur was a dangerous business. Modern anteaters have
hooklike claws on their forefeet, while the most dangerous mod-
ern ground bird, the cassowary (a two-hundred-pound flightless
creature from New Guinea), has a big hind claw. Together, ant-
eaters and cassowaries demonstrate how anchisaurs fought. Living
species of anteaters grow only up to 150 pounds maximum weight,
but their hooked foreclaws are potent weapons of defense against
DEFENSE WITHOUT ARMOR
255
The first dinosaur panzer— Scelidosaurus. Early in the Jurassic Period, the
scelidosaur clan evolved top and side armor composed of stout bony cones.
The twenty-foot-long predator Dilophosaurus would have found it hard to
deliver an effective bite against this defense (dilophosaurs had two thin bony
crests running down their snout—probably a sexual advertising device).
jaguars. When angry and cornered, the anteater stands erect on its
hind feet and tail and lashes out with left and right swings of its
foreclaws. Knowledgeable zoologists take great care in the face of
this attack, for if the anteater strikes full force in a vulnerable area,
such as the stomach, its great claws can effect a full disembowel-
ment.
Anchisaurs' tails were stoutly muscled and they could easily
have reared up, foreclaws at the ready, to face their enemies. An-
256 I DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
chisaur hind claws, especially the one located on the large inner
toe, could lash out with even more powerful blows than the
foreclaws. Cassowaries jump to strike with the full force of their
massive thighs behind their long inner toe. Zoo keepers always
treat cassowaries with the utmost respect—these birds are much
more dangerous than their bigger cousin, the ostrich. And they
are just plain mean, often attacking humans without provocation.
Yet anchisaurs grew to much larger sizes than do cassowaries: a
half-ton anchisaur could have unleashed a kick five times more
powerful than can any cassowary.
The predators that threatened to attack throughout the Juras-
sic and Cretaceous Periods were the long-legged theropod dino-
Long-snouted Coelophysis
attacks an anchisaur.
DEFENSE WITHOUT ARMOR | 257
saurs. This clan is best known for Allosaurus of the late Jurassic
Period and for Tyrannosaurus belonging to the last days of the
Cretaceous. The very earliest theropod meat-eaters had appeared
side by side with the anchisaurs and the other early species of di-
nosaurs during the Late Triassic.
From the earliest days of dinosaur hunting in the mid-1800s,
these predatory dinosaurs, especially those from the Triassic, have
constituted the most cherished discoveries of any field expedition.
The reason is simple—they are quite rare. Over six seasons in the
field digging for dinosaurs, I have personally seen only one pred-
ator skull, one battered predator backbone, and one predator claw.
On average, in the Jurassic beds, one can't expect more than one
Allosaurus skeleton at most per ten brontosaurs. This scarcity of
predator remains is especially acute for the dawn of the Age of
Dinosaurs, the end of the Triassic and beginning of the Jurassic.
For over a hundred years paleontologists sought predator skele-
tons from this earliest epoch with disappointing results.
But two great discoveries during the last thirty years have
provided us with a wonderful glimpse of the first predatory dino-
saurs. The first was the grandest of all: not just one perfect skull,
nor one complete skeleton, but a whole quarry filled with the
complete and partial skeletons of one Late Triassic species, all
preserved in the red mudstone of Ghost Ranch, New Mexico. Ned
Colbert of the American Museum made this discovery, and under
his direction the museum technicians have erected quite beautiful
to dozens of institutions throughout the international community
of scholars. Colbert had stumbled upon a most unusual prize: a
predator trap, a pocket of mudstone that formed in a peculiar lo-
cale where predators had huddled together in death. Predator traps
constitute one of the most puzzling enigmas in paleontology. What
would have attracted meat-eaters to one small spot a few hundred
yards wide, and what had killed and buried them there?
The Tar Pits at La Brea, California, dating from a time late in
the Age of Mammals, are the best-known predator traps and the
best-studied. And they shed some light on Colbert's Triassic pred-
ator trap. La Brea is filled with saber-tooth cats and huge wolves,
all jumbled together in tar-soaked sand which dates from about
twenty thousand years ago. A few plant-eaters—mammoths, cam-
258 I DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
els, horses, and others—have been quarried out of La Brea, but
the overwhelming majority of bones are those of the big meat-
eaters. La Brea seems to have been a deathtrap for wolves and big
cats, acting much like a giant sticky flypaper surface whose tar-
soaked sand entrapped the meat-eating mammals' paws in viscous
asphalt, miring them until the exhausted beasts sank down and died.
Dead and dying animals would attract more predators to the tar
sweeps, unwary meat-eaters who thought they could get a meal
with little effort. And each new victim would add to the lure.
Could Ghost Ranch have been such a flypaper trap? The site
hasn't yet been analyzed sufficiently to yield any conclusions. The
mudstone at Ghost Ranch did not yield the slightest trace of as-
phalt. But it's not impossible that sticky mud might have served
to have the same effect, trapping dinosaur feet in a viscous, ines-
capable mire.
Colbert's splendid skeletons seem to belong to the same ge-
nus Cope had named from fragments in 1880: Coelophysis, roughly
translated as "hollow-boned beast." Hollow it indeed was—all of
the major limb bones and vertebrae were constructed like those
of birds, with an outer shell of dense bone rind surrounding an
empty core. So perfect are Colbert's skeletons that no guesswork
is required to reconstruct these bodies. Coelophysis was small com-
pared to its Jurassic nephews Allosaurus and Ceratosaurus; the fully
adult length was only six feet, half of which was tail. Compared to
those Jurassic predators, Coelophysis was long and slender in the
torso and very long in the neck—the neck, body, and tail all seem
to flow into one another to create an unusually smooth profile. Al-
though it appeared early in dinosaur history, Coelophysis was al-
ready a birdlike biped with wide upper hip bones and deep lower
hip bones, the whole design providing for ample thigh muscles and
quick thrusts of the hind leg. The vertebrae in the neck were an-
gled, producing a natural S-shaped curve, so the head was carried
high above the shoulders as a bird's would be.
How did Coelophysis hunt? Its graceful yet strong neck could
lunge forward for a quick snap at a small prey or for hit-and-run
attacks against large prey. Teeth always provide the best biome-
chanical clues to the killing tactics. Coelophysis's dental pattern was
totally different from the killing apparatus we find in mammalian
predators—wolves, leopards, and lions. When a wolf or cat bites,
DEFENSE WITHOUT ARMOR | 259
the four fanglike front teeth (canines) penetrate deeply into the
prey. This bite is precise. The lower canine pair bites just in front
of the upper, and the two pairs together lock the prey in a killing
grip. Cats and foxes kill rabbits with one bite through the nape of
the neck. Lions kill big prey—zebras and buffalo—by clamping their
canines down on the throat and holding on until the prey suffo-
cates. Coelophysis's killing teeth were organized for a very different
technique. Instead of two pairs of canine fangs, Coelophysis had a
long row of small, curved, daggerlike teeth, each with the sharp,
serrated edge both fore and aft characteristic of nearly all hunting
dinosaurs. A bite from such an assemblage of teeth would have
left a long, shallow wound across the prey's flesh.
Coelophysis's teeth were designed to slash through flesh, not
to hold it. A cross section of this dinosaur's tooth shows a tear-
drop outline, with a blunt, rounded front edge and a tapered,
sharply chiseled rear. When Coelophysis bit through its prey's hide,
the blunt front edge prevented the prey from slipping away while
Elegantly designed meat
slicer—the skull of
Coelophysis. This lively Early
Jurassic hunter epitomizes
the light, flexible
construction of paper-thin
bony sheets and slender
struts.
260 I DEFENSE, LOCOMOTION, AND THE CASE FOR WARM-BLOODED DINOSAURS
the sharp rear edge slashed through the flesh. The backwardly
curved tips of the teeth assisted in driving the whole tooth row
backward through the wound. Saw-toothed from top to bottom,
the serrated rear edges were designed to assist this backwardly di-
rected slash, since they would allow the entire tooth to saw back-
ward through hide and muscle. But on the front edge the serration
was only at the tip. The blunt base along most of the front of the
tooth was smooth, and so would hold the prey as it struggled to
free itself. All of the structural details were cunningly calculated
to permit the tooth to act as both knife and fork, cutting and
holding.
Some living species of monitor lizards have teeth like those
of Coelophysis, and these lizards inflict long, jagged wounds when
they bite. Komodo dragons, the biggest monitor lizards alive to-
day, can even kill cows and people with the wounds they inflict.
Since both monitor and dinosaur teeth curve backward, the jaw
muscles must be arranged to pull the teeth rearward as the jaws
Coelophysis teeth worked like a combination fork
and steak knife. All the teeth were backwardly
curved blades with saw edges running along the
entire trailing edge and the tip of the leading edge.
In cross section (shown in black) the leading edge
was blunt but the trailing edge was very sharp.
Upper teeth were much larger and much sharper
along their trailing edges than lower teeth. But
lower teeth had stronger, blunter leading edges. So
when the biting muscles contracted, the lower teeth
held the prey and prevented it from slipping out of
the mouth while the sharp upper teeth slashed
backward, making a long, nasty wound.
DEFENSE WITHOUT ARMOR
261
Dinosaurian answer to the electric
carving knife— Coelophysis biting
mechanism. The huge hole in the
snout bones was for the big muscle
that powered the bite. A much
smaller biting muscle was in the
hole just behind the eye socket, and
the jaw-opening muscle was strung
from prongs sticking backward from
the head and jaws. The big snout
muscle was arranged to pull the
skull down and backward (line of
pull shown by the black arrow). So
when the muscle contracted, the big
upper teeth slashed back and
downward toward the lower teeth.
close. The skull and jaws of the lizards feature extra joints rather
like a snake's to permit this. Yet Coelophysis's killing bite must have
been different from the lizards' in one fundamental way. The up-
per and lower teeth of Komodo dragons are the same size, but in
Coelophysis the upper teeth were much larger than the lower. Con-
sequently, more muscle power was required to pull the upper teeth
back through prey than was necessary for the lower. The upper
teeth were also more sharply edged, so they must have produced
more of the cutting action, while the lowers did more of the hold-