Covering The Body

Skin is the first body system that we notice in a living animal, but it is also the first part of a buried animal to disappear. Even though the skin in a large animal may form a tough shield several centimeters thick in places, it usually rots away eventually after death. For some Ice Age mammals, such as giant ground sloths, wooly mammoths, or bison, skin with its hair is, under special conditions, actually preserved, but these fossils are at most only a few tens of thousands of years old.3 Regrettably, this is not the case for dinosaurs; no actual skin survives. Was ceratopsian skin furred, feathered, bare, or scaled? Fortunately, in some cases, the carcass of an animal may make an impression in finegrained mud before the ravages of decay proceed. That is the case with the Canadian horned dinosaur Chasmosaurus, for which an impression of a patch of skin was preserved with a skeletal specimen.4 The skin, which comes from the pelvic region, might be best described as reptilian in nature. Large circular plates up to 55 mm in diameter are set in irregular rows at a spacing of 50 mm from each other. Between the large round plates are irregular polygonal plates a centimeter or less across. Unfortunately neither Chasmosaurus nor any other ceratopsian shows the pattern of skin over the skull or frill.

There is one other matter concerning the body covering or integument, and that has to do with the covering of the horns. Living bare bone is not exposed to the environment.5 It is in point of fact an inference that horned dinosaurs had horn—not a risky inference, but a real one nonetheless. That is because horn refers to a specific anatomical tissue that is never preserved. Horn is composed of a protein substance called keratin (Greek: made of horn). What we call horns on the skulls of horned dinosaurs are in fact horn cores. The true horn is a keratinous sheath that covers the bony horn core. In modern horn-bearing mammals, the bovids (such as sheep, cattle, and antelope), the horn sheath extends a variable, and often considerable, distance beyond the tip of the horn core. Horn is nonliving material; it is insensitive to pain and lacks blood supply. The horn core underneath is certainly living bone, and it is also covered by a deep layer of skin called dermis, the site of the blood and nerves that indirectly provide sensitivity to the horn. New horn is added around the base, and old horn may be worn off the tip by rubbing, a behavior that is common among horn-bearing animals. I am certain that the horn cores of horned dinosaurs were covered with true horn to protect the living bone underneath. Ceratopsian horn cores show grooves and channels that suggest that a blood supply ran underneath the horn, as expected.

Not only farmers are familiar with horn: all of us experience horn on our fingers and toes. Nails are essentially horn, that is, keratin. We are familiar with the flexibility, insensitivity, and growth of our nails. Nails are specialized flattened keratin structures covering the ungual bones (the terminal bones of the hands and feet) of primates. Claws are more generalized structures covering the pointed ungual phalanges among a

Physiologus Adler
FIG. 2.1. Fleshed-out reconstruction of Chasmosaurus belli. (Robert Walters.)

wide range of vertebrates, including reptiles, both herbivorous and carnivorous. It is clear enough that many dinosaurs had claws that were similar to those of living claw-bearers, whether reptiles or mammals. Horned dinosaurs did not have claws; the unguals instead were broad and rounded. To me they look like nothing so much as a horse ungual. I do not think, however, that there was an exposed hoof as in horses. The feet of ceratopsians, like those of virtually all dinosaurs, were digitigrade. That is to say, the toes were flat on the ground, and were for the most part invested in dense skin. I believe the unguals of horned dinosaurs were covered in a somewhat hoof-like covering on both the upper and lower surfaces. As I sit here at my desk writing, I am examining the ungual of what may be a Chasmosaurus (although one cannot tell for sure from an isolated bone, a "spare part"). It is coffee-brown and heavy, unlike the light, ivory-colored horse ungual beside it. But the two bones share some striking similarities. In both, I see a pair of holes near the proximal articular surface where the arteries that run down the leg and continue onto the toe entered the bone to form a terminal arterial arch. And along the crescentic tapered edge I see a series of holes where blood from the terminal arch left the spongy bone of the toe bone to nourish the dermis underneath the hoof. In fact, I see more similarities than differences between the ungual of the horse and the ungual of the horned dinosaur.

The final stage of anatomy is to answer the question: what did the animal look like? In a real sense this is not the role of the scientist but of the artist (Fig. 2.1). Twice blessed is the scientist who is also an artist, for example, David B. Weishampel of the Johns Hopkins University; Dave was my student. There are also artists who contribute to the scientific literature, notably Gregory Paul.' For most of us paleontologists, restoring the appearance of the animal is beyond our skills. Our work typically ceases with description and interpretation of skeletal remains. Often the bones then repose in the tranquillity of a museum drawer. Sometimes there is sufficient interest that they are then mounted by highly skilled museum technicians, who also possess skills that many of us scientists lack. It may also come to pass that we have the opportunity to work with an artist to restore the appearance of our favorite animal. Such collaboration between scientist and artist has long been fruitful. When Richard Owen coined the name dinosaur in 1842, it had no impact on the public—none whatsoever. But twelve years later, when he teamed up with the eccentric sculptor and artist Benjamin Waterhouse Hawkins, the dinosaur at last took tangible form and became accessible to the Victorian public—and dinomania was born!

THE SKELETON: FANTASIA ON A THEME

The major body system the paleontologist works with day in and day out is the skeletal system, the bones. It is to the skeleton that we must now turn our attention. Knowledge of the skeletal system forms the great divide between those who know a few dinosaur names and can recognize Triceratops in a museum and those who are really eager to learn paleontology. (I still remember the thrill I felt on Christmas Day many years ago when the Tyrannosaurus model my parents gave me came with the bones named—I really felt like a paleontologist then!)

I have always wanted a ceratopsian skeleton in my living room. I finally found one, a high-quality epoxy resin cast of Chasmosaurus belli,7 in a catalog. The price seemed almost too good to be true, so I quickly ordered one from a mail-order house that I had never heard of: Big Bob's Bargain Bone Barn. In due course a large and rickety crate appears on my doorstep. I barely manage to get it inside when the box collapses and an unbelievable cascade of bones rumbles across my living room floor—more than three hundred of them in all sizes and shapes! I look in vain amid the chaos for assembly instructions or a key for identifying the bones. I indignantly call Big Bob's number, only to find it disconnected—Big Bob is no more. It was too good to be true, or at least too good to last. What to do?

I decide not to panic. Although the number of bones in a Chasmosaurus skeleton is large, it is finite, and many shapes are very distinctive,

Chasmosaurus

FIG. 2.2. Exploded skeleton of Chasmosaurus. The text describes the magnificent Chasmosaurus belli skeletons at the Canadian Museum of Nature in Ottawa, but the skull shown is that of Chasmosaurus mariscalensis from Texas, which better serves to illustrate the separate skull elements. (Bruce J. Mohn/Robert Walters.)

FIG. 2.2. Exploded skeleton of Chasmosaurus. The text describes the magnificent Chasmosaurus belli skeletons at the Canadian Museum of Nature in Ottawa, but the skull shown is that of Chasmosaurus mariscalensis from Texas, which better serves to illustrate the separate skull elements. (Bruce J. Mohn/Robert Walters.)

so I decide to try to sort them into groups according to their shapes. There are still several hours left before my wife is to return home. I also have a couple of good books in my home library. The best one for this task is a thick and daunting technical treatise called The Dinosauria, which has detailed chapters on each group of dinosaurs.'! turn to the

FIG. 2.3. Representative vertebrae of Triceratops. (a) Syncervical or fused vertebrae of the neck. The first vertebra is only a small vestige, but the second, third, and fourth vertebrae are completely and inseparably joined. In this specimen, the fifth vertebra is joined by sediment but is not part of the syncervical. (b and c) Second dorsal vertebra in side view and front view, (d and e) First caudal vertebra in side and front views, (f, g, and h) Distal caudal vertebra in side, front, and bottom views. The parts of the vertebrae are labeled as follows: a, cranial articular surface of the vertebral centrum or body; p, cranial articular surface of the vertebral centrum or body; hypophysis (h) and diapophysis (t) are facets for attachment of the rib head and tubercle, respectively; t also denotes the transverse process; n, nerve or spinal canal; r, rib; s, spinous process; z and z', cranial and caudal articular processes, respectively. (From Marsh 1'91a.)

Caudal Articular Processes

last chapter of the book, entitled "Neoceratopsia."' It looks like it will be a useful guide. So I sit on the floor with an ethereal smile on my face and sort the bones into piles: big bones and little ones, limb bones and vertebrae, ribs and toe bones, and just plain odd bones (Fig. 2.2).

Vertebrae are easy to recognize. They have large spool-shaped bodies, with blade-like neural spines on top, enclosing a hollow canal where the spinal cord once sat. The vertebrae differ in size, but I assemble a pile of sixty-three separate vertebrae. There are also two units where the spools are joined together and cannot be separated. One consists of what seem to be three vertebrae, with a deep cup at the end. The other is a long unit with ten spools joined together, with flat faces at both ends. I learn that the first, called the syncervical, which is 28 cm long,10 actually represents a joining or fusion of the first four neck or cervical vertebrae that helped support the heavy head (Fig. 2.3a). The deep cup on the face of the first vertebra forms a socket in which the ball at the back of the skull sat. The second set of fused vertebrae is called the sacrum. It is 72 cm long and represents the connection of the vertebral column to the pelvis.

I am able to arrange the vertebrae between the syncervical and the sacrum into a series in which the bodies, bony processes, and canal change gradually from one to the next. The separate vertebrae of the neck region, six in number, have round faces about 12 cm wide; the canal for the spinal cord is fairly wide; and the neural spines are rather low. The next region of the body is called the dorsal region, which includes the thoracic and lumbar regions of mammals. Most dinosaurs did not have a rib-free region in front of the hips, which is called the lumbar region in mammals. Twelve vertebrae belong to the dorsal region. Here the neural spines grow tall, and the transverse processes become prominent, swept-up wings to which the ribs attach (Figs. 2.3b,c). At their tallest in the mid-dorsal region, the vertebrae are about 40 cm tall. At the front, the vertebrae are rather erect, but farther back they have a backward lean to them, and the neural canal for the spinal cord is decidedly narrower than it was farther forward. The vertebrae have narrowed to about 10 cm in width. The sacrum is long and heavy, and the neural spines and transverse processes form almost continuous surfaces at right angles to each other (see Fig. 2.5b). In addition, heavy specialized ribs form a bar for attachment to the bony pelvis. There are forty-five tail or caudal vertebrae. They start at the base of the tail as tall, erect vertebrae with long spines and strong horizontal transverse processes, but with short bodies (Fig. 2.3d). Proceeding backward, the spines quickly decrease in height, and the transverse processes dwindle and disappear halfway down the tail. Toward the end of the tail the vertebrae are simple spools, longer than they are high (Fig. 2.3e).

I feel pleased with the results of this sort. As I join the vertebrae together with their interlocking articular processes (zygapophyses) and lay them out on my floor, they stretch 4.1 m in length, including 1.5 m from the sacrum to the front of the spinal column and 1.5 m from the back of the sacrum to the tip of the tail.

Ribs also seem reasonable to deal with. I have a pile of forty-two bones or twenty-one pairs, many of which I am certain are ribs, since they are up to 75 cm long, thin, and curved like the staves of a barrel. Others, however, are short, only 15 cm long, straight, and forked at one end like a slingshot, or like an asymmetrical lowercase y. In between are ribs of intermediate shapes and lengths. In fact, they all fit together into a continuous series, with no sudden jumps in size between them. This makes me certain that they are all ribs. The short slingshot-shaped ones attach perfectly to the vertebrae at the front end of the neck. In fact, there are clear points of attachment; one fork of the slingshot attaches to the side of the spool, the other to the transverse process. When attached to the vertebrae, the neck ribs point backward. The neck ribs closest to the chest increase in length and, though still straight, no longer look like slingshots at their upper ends. The first several chest or dorsal ribs are T-shaped, and both knobs at the upper end of the rib attach to the transverse process of the vertebrae; the ribs point down, not backward. Most chest ribs are long and become curved like barrel hoops. The last ribs next to the pelvis are shorter than the typical dorsal ribs (Fig. 2.4).

Another group of bones seems to be related to the vertebral column in some way. They too are slingshot-shaped, but quite symmetrical, which suggests to me they are situated on the midline of the animal, and therefore single rather than paired. There are thirty-two of them, the longest about 75 mm long and the smallest a quarter of that size. I determine that they are the chevron bones or hemal arches. They hang beneath the tail, the counterpart of the neural spines above the tail vertebrae. The tail vertebrae show facets on their undersurfaces where the chevrons attach. The canal framed by the slingshot encloses blood vessels to the tail, accounting for the descriptor hemal, which refers to blood.

Now I turn my attention to the long bones. I make a pile of twenty bones. There are also four others that are only 30 cm or so long. They too might belong with the long bones. Some of them I am certain are leg

Brachyceratops
FIG. 2.4. Representative ribs from Brachyceratops. (a) Second dorsal rib; (b) a middle dorsal rib; (c) another middle dorsal rib; (d) final dorsal rib. (From Gilmore 1917.)

bones, but others are either flat and broad or curved, so I am not quite so certain, except they are as long as the bones about which I am certain.

I can make a pretty good guess at the pelvic bones. I know that there are three on each side, the ilium, the ischium, and the pubis. The ilium is dorsal; the ischium and pubis are the two ventral bones.11 Mammals, birds, crocodiles—in fact all land vertebrates—have the same three hip bones. The ilium is a long, somewhat flat bone. At 96 cm in length, no other bone in the skeleton equals it in length. It has a blade that projects forward (cranially), another that projects backward (caudally), and a shallow notch between them that faces ventrally and defines part of the hip joint. The cranial blade lies horizontally and forms an overhanging shelf above the hip. On the internal (medial) surface of the ilium are a set of ten scars that show the points of attachment of the ten sacral vertebrae to the ilium. The ischium is very distinctive in horned dinosaurs, because its long, simple shaft points backward underneath the tail and curves downward. It measures 70 cm in length. The forked upper end touches the ilium above and the pubis below. The arc in between defines more of the hip joint. The hip joint is completed by the third bone, the pubis, on the cranioventral aspect of the pelvis. The pubis consists primarily of a vertical blade about 45 cm long, which spreads laterally from the hip joint toward the ribs. It has a small rod caudally that parallels the ischium for a short distance (Fig. 2.5).

The longest pair of columnar straight bones measures 75 cm in length. These have to be the thighbones or femora. Each femur has a heavy, ball-shaped head leaning inward, and at the lower end a pair of curved surfaces or condyles that form a roller surface at the knee joint. There is a prominent depression with a bony tab next to it on the inside surface of the middle of the femur, which is the site of attachment of the major muscle, the caudifemoralis, which we have already met. Also distinctive is a pair of somewhat shorter but robust bones that measure 53 cm in length. These are the tibiae or shinbones. Each tibia reminds me of a bowtie, being broad at the upper and lower ends and forming a slender shaft in the middle. The expanded upper end points forward and forms a prominence at the knee, whereas the lower end is expanded side to side and forms a hinge for the ankle joint. Each tibia also has a long slender rod of bone beside it, called the fibula. I now have a good idea of the size of the hindlimb of Chasmosaurus, because I know the foot is not going to add that much to its length. I am struck by the thought that because the shin is so much shorter than the thigh, these dinosaurs must not have been terribly swift runners (Fig. 2.6).

Before I turn to the feet, I decide to work on the front legs. There is a pair of long flat bones with a slight curve to them. These are not leg bones, but they must have something to do with the legs. At 6' cm in length, each bone is longer than any of the true leg bones of the front limb. I recognize it as the scapula or shoulder blade, quite similar in form to that of a chicken. It is gently curved inward to conform to the shape of the rib cage. The lower end has an attachment surface for another bone, the dinner-plate-sized coracoid. The lower end of the coracoid lies near the sternum or breastbone, which consists of a pair of flat, kidney bean-shaped plates that sit on either side of the midline of

Kidneys Sit Back Scapula

the body on the lower or ventral surface of the chest. This arrangement is totally unlike that of mammals. A prominent notch shared by the scapula and the coracoid on the caudal side represents the shoulder joint. The biggest of the remaining bones is the humerus, the bone of the upper arm, which is about 50 cm long. It has a bulbous head and a wide blade that extends halfway down the shaft before it narrows, then expands modestly at the elbow. As in the hind leg, there are two bones for the middle segment of the front leg. The larger of the two is the ulna, which is 43 cm long and has a prominent process on its upper end. This process corresponds to the point of our elbow, the funny bone. The other bone is the radius, about 32 cm long, which is a rather nondescript shaft (Fig. 2.7).

The job has been fairly straightforward up to now. I have been avoiding dealing with the foot bones because they are the smallest bones, and there are so many of them. I have a collection of ninety-two bones, which range in length from shafts of 21 cm to little nubbins 13 mm long.12 After lengthy contemplation I manage to sort them into two groups: a larger, more robust set of forty-eight bones representing the hind feet and a slightly shorter, more lightly built set of forty-four bones representing the forefeet. I work on the front feet first. The longest bones of the front foot are the five metacarpals (which correspond to the knuckle bones of our hand), two of which are about 13 cm long and the other three of which are 8-10 cm long. They are labeled metacarpals I-V, with Roman numerals usually designating the digits (fingers) from medial to lateral (that is, from thumb to pinkie). Metacarpal (MC) III is the largest, corresponding to the central axis of the forefoot. It is barely longer than MC II but is much wider at its proximal end.11 MC II is barely shorter but more slender. Metacarpals I and IV are both robust, but MC I is 2 cm shorter than MC IV. MC V is much less substantial than the other four.

I am quite confident about assembling the metacarpals and feel proud of what I have done. The phalanges are quite another matter. I find it very useful to have prior experience with the pattern of the

FIG. 2.5. Pelvis of Triceratops in (a) left lateral view and (b) dorsal view. In (a), the ilium (il) is dorsal, the pubis (p) points down and forward, and the ischium (is) curves down and back. All three hip bones meet at the hip joint or acetabu-lum (a). In (b), the broad, shelf-like ilia (il) join the elaborate fused sacrum. (From Marsh 18'1b.)

Dinosaur Tibia

FlG. 2.6. Ceratopsian left hind limb, cranial view. The femur and tibia are of Triceratops; the foot is based on Centrosaurus. a, astragalus; h, head of femur; t, greater trochanter. (From Marsh 1891b and Brown 1917.)

Triceratops Scapula
FIG. 2.7. Ceratopsian left forelimb, cranial view. Scapula, humerus, and ulna based on Triceratops; radius and manus based on Centrosaurus. (From Marsh 1891b and Lull 1933.)

digits. I read that the ceratopsian hand, or forefoot if you prefer (manus, the anatomical term, covers both designations), departs a little from the norm, having fourteen phalanges arranged in a formula of 2-3-4-3-2, which indicates that the outer digits are somewhat reduced in size.14 The most obvious feature of the phalanges are the unguals, the terminal phalanges or hoof bones. These are the flattened, crescent-shaped last segments of the first three toes. The unguals of horned dinosaurs are quite similar to those of hadrosaurs or duck-billed dinosaurs, and even similar to those of modern horses (except that horses have only one on each foot). This suggests to me that horned dinosaurs had some sort of specialized hoof-like structure enclosing both the upper and lower surfaces of each toe. The outer two toes end in little nubbins of bone. Otherwise the remaining eighteen phalanges are not very distinctive. I have no chance of separating left from right. The first or proximal row of phalanges are a little longer than the more distal ones; those of digits II and III are wider than the others. I try my best to make each foot look like the picture in the book. I also place three small flattened disks, the carpal or wrist bones, at the upper ends of the metacarpal bones. They are smaller than the surfaces of the metacarpals, so no precise fit is possible. It seems that there were cartilage, dense connective tissue, and tendons, all of which contribute to a definitive fit. As these materials do not fossilize, we are left with a bit of a gap.

My work still isn't perfect, but it is the best I can accomplish with a jumble of bones. It is much nicer when an entire skeleton is found completely articulated, that is, in life position, with no guessing required. But I think you would be impressed by my results. Now the hind feet do not seem so daunting. The metatarsals are longer and heavier than the metacarpals that they otherwise resemble. There are four major ones that range between 12 and 21 cm in length. A fifth metatarsal (MT V) we describe as vestigial: at less than ' cm in length, it is a very much reduced and somewhat useless version of a functional metatarsal. It has no toe bones associated with it, and, like the splint bones of a horse leg, would not have been visible in a living animal. As in the forefoot, the central bone, in this case MT III, is the longest, and the second, MT II, is next. MT IV is robust but 5.5 cm shorter than MT III. MT I, at 12 cm, is slightly shorter but still more robust than the longer metacarpals. The phalangeal formula of the foot is 2-3-4-5-0. This means we only have to worry about phalanges for four toes instead of five. They are all wider than the manual phalanges (i.e., those of the manus or hand), and, as in the hand, the proximal row of phalanges is longer than any of the others. Each of the four digits bears a full ungual phalanx; there are no nubbins in the foot. One characteristic of the fourth digit is that because it has the most phalanges (five), it follows that each of these is relatively short; in fact, the lengths of all the nonungual phalanges decrease from the first digit to the fourth digit. This is true not just of horned dinosaurs or even of dinosaurs, but also of birds and all vertebrates that have a phalangeal formula of 2-3-4-5. These characteristics are sufficient to allow me to put the feet together pretty convincingly.

I have a few other bones left over. These are the tarsal bones. The largest one, the astragalus, is 18 cm wide and forms a roller surface that caps the lower end of the tibia and constitutes a major component of the ankle joint.15 The smaller, block-like calcaneum extends the ankle joint surface laterally over to the lower end of the fibula. There are three irregular, flattened disks constituting the distal tarsals that in life were probably attached by bands of connective tissue to the proximal ends of the metatarsals. I can only guess at their exact positions.

I am now the proud possessor of a semiarticulated Chasmosaurus skeleton. It measures 4.1 m (13 ft 6 in.) in length without the skull, laid out along my living room floor and snaking into the dining room. When I prop the skeleton16 up in a quasi-reasonable posture, it measures 1.5 m (5 ft) high at the hips. How is it that I feel that it is not going to be a permanent fixture in my living room? If I had a drawing room, it might go there. Sadly, however, I do not live such a life-style. I know my daughter would helpfully suggest I put it in the freezer, beside the frozen alligator legs—teenagers can be so cheeky! Maybe it will have to go into the garage. I refuse to consider the backyard. Anyway, my wife has long since come and gone. At least she said nothing instead of what she really thought. It is past midnight. I have identified and positioned 255 bones. I am exhausted. The skull bones can wait.

The next day I am bright-eyed and bushy-tailed, eager to get back at it. Again I am confronted with a jumble of bones, albeit a smaller pile than before. These skull bones are much more irregular than the bones that form the skeleton—not much by way of cylinders, tubes, or rods here. Some have conspicuous holes, grooves, or channels that probably permitted nerves or blood vessels to pass through. Some have interesting textures or patterns on the surface. Most bones of the skull come in pairs, with the bone on the right side of the animal being a mirror image of the one on the left.17 Some bones on the midline of the body are single and symmetrical. The longest bone in the pile is a whopping 87 cm

Triceratops Epijugal

coronoid deritary

FIG. 2.8. Exploded skull of Chasmosaurusmariscalensis, University of Texas at El Paso, viewed from the right side. (After Lehman 1989. Bruce J. Mohn/Robert Walters.)

long, longer than the longest limb bone. Can it really be a skull bone? Its symmetrical shape suggests it belongs on the midline, and the fairly slender shape of the processes shows that it could not have borne the weight of the body as the limbs must do. The smallest bones are just a couple of centimeters long. It might be tough solving the whole puzzle, but certain bones stand out. The four jaws with teeth are obvious enough, two uppers and two lowers. To lay them out is an obvious starting point. Also obvious are the two bones that form the beak at the front of the skull, and three bones that bear horns on the top of the skull. A good guess is that the long T-shaped bone and a pair of other, somewhat fancier bones, only a little shorter, belong to the crest that sticks out behind the skull. Laying out these few bones gives a very useful but rough shape to the skull. Now I can open my book and get down to work (Fig. 2.8).

The upper jaws are called the maxillae (singular maxilla). Each is rather triangular in form, with a long, straight base corresponding to

FIG. 2.9. Upper or maxillary tooth of Triceratops. (a) Lateral view, showing enamel ridge; (b) side view, showing split root. (From Marsh 1891a.)

Tooth Side View With Roots

the position of the tooth row, and with the apex pointing dorsally. The maxilla is about 36 cm long and half that high. Twenty-eight teeth can be counted (Fig. 2.9). The bone in front of the maxilla is called the pre-maxilla—the logic for this name being refreshingly clear. It is 36 cm long and 22 cm high. The paired premaxillae are half round and frame a large part of the opening for the nostrils (external nares), which are usually large in horned dinosaurs. In front, the two bones are in broad contact with each other, but behind the two bones diverge somewhat. A rod of premaxilla ascends along the front surface of the maxilla. There is a distinctive dimple or pit in the bone in front of the nostril, and a tab of bone projects into the floor of the nostril. Horned dinosaurs had a turtle-like toothless beak composed of two bones, the rostral bone dor-sally and the predentary ventrally. Both bones are robust, have sharp keels along the midline, end in points, and have sloping cutting edges; the edge on the rostral bone continues backward along the ventral edge of the premaxilla. The rostral and predentary bones each have a pronounced texture that suggests that they were covered in life with a horny material such as that which covers the beak of a turtle or a bird. The rostral bone is found only in horned dinosaurs, but the predentary is found in all the ornithischian or bird-hipped dinosaurs.

The nasal bone is easy to recognize. It is a good-sized, saddle-shaped bone 29 cm long that covers the external nostril and bears the horn core over the nose, the nasal horn core. The horn core is rather low and blunt. At only 10 cm in height above the bridge of the nose, it really could not have been too menacing. Behind it, the nasal bone sits on the premaxilla and maxilla, forming both the roof of the nasal cavity inside and the bridge of bone between the eyes and the nose horn. The orbital horns over the eyes are well preserved but, interestingly enough, are of rather different heights, the left one being 14 cm high over the eye socket, the right one only ' cm high. The left horn core ends in a somewhat blunt point and curves backward (it is said to be recurved). The right horn core happens to be much blunter. Again, neither horn core seems particularly threatening. Each horn is an outgrowth of the postorbital bone, an extensive element that forms much of the upper and back borders of the eye socket or orbit. A variety of names are used to designate the horns over the eyes. Among them are "brow horn," "supraorbital horn," and "postorbital horn." They refer, respectively, to the brows, the position of the horn above the eyes, and the bone that composes the major portion.

I am pleased that each postorbital has some other, smaller bones fused to it so that I do not have to struggle with them in order to fit them into the whole skull. These smaller bones, including the lacrimal bone (the tear bone, because it contains the tear duct), the prefrontal, and the supraorbital, are hard to fit. Collectively these bones can be termed the circumorbital series, because they surround the orbit. Another cir-cumorbital bone forms a major component of the cheek as well. This bone is called the jugal. In Qiasmosaurus it measures 27 cm by 27 cm, and it is anything but rectangular. It is notched above, where it forms the lower rim of the orbit, and ends below in a tapering point. A second notch along the rear border defines a portion of the opening called the infratemporal fenestra.1' In some horned dinosaurs, especially Penta-ceratops, the point is so thick and prominent, accentuated by a small bone termed the epijugal ("on top of the jugal") that may fuse to the tip of the jugal, that it is described as a jugal "horn" (hence the name Pentaceratops, which means "five-horned face," instead of the usual three-horned-faced design of most ceratopsids).

The quadrate bones are complex and distinctive bones that are braced on the back of the skull to support the lower jaws.19 In lateral (side) view of the skull each quadrate is somewhat covered by the jugal, and it is seen as a sloping rod that is expanded transversely at the lower end, where it attaches to the lower jaw. The upper end is a surprisingly thin blade, and it also has a wing-like process that attaches to delicate bones of the palate (roof of the mouth). The bone measures about 20 cm in length, and the expanded lower end is about ' cm wide. When the skull is viewed from behind, the quadrate bones are prominent vertical props just inside and underneath the frill. The quadrate does not actually touch the jugal, but a spacer element is wedged between the quadrate and the inner surface of the jugal. This element is called the quadrato-

jugal, named for the two bones it separates. It is more than 3 cm thick ventrally but thins dorsally as it wraps around the shaft of the quadrate.

The bones of the frill are unmistakable.20 There are only three bones, one parietal and two squamosals, making up the bony frill, which projects behind the skull and overhangs the neck. It is a terribly distinctive, extravagant structure. The frill of Chasmosaurus has huge, paired open spaces in it, called parietal fenestrate, measuring about 60 cm in length by 30 cm in width. The central bone of the frill is called the parietal bone. It runs from the postorbitals to the back of the skull. It measures 87 cm in length and has something of a T shape. Its striking features are a central bar 4 or 5 cm thick and a symmetrical crosspiece measuring about 1 m from end to end. Beside the front of the bar, a thin shell of flat bone slopes down toward the supratemporalfenestrae. Adjacent to the supratemporal fenestrae the surface is very smooth. A well-developed channel leads from the fenestra toward the frill. I infer that jaw-closing muscles exited the inside of the skull through the fenestrae and occupied part of the frill, but not necessarily the whole thing.

Besides the parietal, a pair of squamosal bones comprise the frill.21 In ceratopsid ancestors, the squamosal bones were small bones only a centimeter or two long at the top corners of the skull that stabilized the quadrate bone and thus indirectly contributed to the support of the lower jaws. In ceratopsids generally there is an exaggeration of the squa-mosal bones, and in Chasmosaurus and its relatives (the chasmo-saurines) the squamosal is greatly elongated behind the real skull. If the skull is considered a bony box enclosing the brain, the nose and mouth cavities, and the jaw-closing muscles, the frill is literally an add-on behind the skull, an extravagant come-on, as we saw in Chapter 1. In Chasmosaurus the squamosal measures 76 cm in length. It is 27 cm wide, thick, and flat and tapers toward the free end of the frill. At the front the squamosal is keyed firmly to the rest of the skull, especially to the postorbital and to the jugal. The infratemporal fenestra is enclosed by the squamosal and the jugal. The squamosal is constricted to a narrow point just behind the infratemporal fenestra and then flares to its widest point and sweeps backward and upward as it tapers to a point. The free edge of the squamosal is festooned with a series of scallops, seven or more in number, which tend to increase in prominence caudally. These ornamentations probably were separate bones, ineptly named epoccipi-tals, that fused to the squamosal in adult animals. The squamosal contacts the parietal along its length and extends clear to the back corner of

FIG. 2.10. (a) Underside of skull of Chasmosaurus belli, showing position of the occipital condyle, at the back end of the "true" skull. The frill is an "add-on" that doubles the length of the skull, (b and c) Occipital condyle and braincase of Centrosaurus: (b) caudal view; (c) lateral view. The foramen magnum (fin) is the opening through which the spinal cord leaves the braincase. The roman numerals indicate the holes by which the cranial nerves leave the braincase. bo, Basioccipital; fo, fossa ovale; lea, opening for internal carotid artery. (From Dodson and Currie 1990. Donna Sloan. Courtesy of the University of California Press.)

FIG. 2.10. (a) Underside of skull of Chasmosaurus belli, showing position of the occipital condyle, at the back end of the "true" skull. The frill is an "add-on" that doubles the length of the skull, (b and c) Occipital condyle and braincase of Centrosaurus: (b) caudal view; (c) lateral view. The foramen magnum (fin) is the opening through which the spinal cord leaves the braincase. The roman numerals indicate the holes by which the cranial nerves leave the braincase. bo, Basioccipital; fo, fossa ovale; lea, opening for internal carotid artery. (From Dodson and Currie 1990. Donna Sloan. Courtesy of the University of California Press.)

Triceratops Braincase

the frill. On the medial or internal surface of the squamosal there are several distinct ridges that provide slots to stabilize the head of the quadrate and the lateral processes of the braincase, but these joints do not seem all that firmly constructed.

Fortunately for my puzzle-building skills, the braincase presents as a single compound structure rather than as its separate constituents. The braincase is a box that encloses the brain, fits within the greater box of the skull, and also attaches the skull to the vertebral column and the muscles of the neck—a tall order for a single structure. What is most evident about the braincase viewed from behind are the large opening at the back, the spherical knob, and the wings or lateral processes. The opening is the foramen magnum (literally the large hole) by which the spinal cord leaves the brain and travels through the body in the neural canal of the vertebrae. The knob is known as the occipital condyle, and it is another highly distinctive feature of ceratopsids (Fig. 2.10).22 No other kind of dinosaur has such a large, perfectly spherical structure that projects so prominently away from the skull. The structure of the con-dyle speaks of a high degree of mobility or maneuverability of the head. The width of the condyle in our specimen is 67 mm, only slightly smaller than a tennis ball. The braincase consists of both unpaired and paired elements, difficult to distinguish in adult specimens. The condyle consists of three bones, including the paired exoccipitals and the unpaired basioccipital, each contributing about one-third to the whole structure. The exoccipitals constitute much of the occipital surface and also form a pair of wing-like or even fan-shaped lateral processes that meet the squamosal and support the upper end of the quadrate. The basioccipital is an unpaired midline bone that lies underneath the brain. Besides contributing to the condyle, it has robust processes for attachment of muscles from the neck. Two pairs of modest openings for nerves are located on each side of the condyle.23 The basioccipital continues forward underneath the brain as the basisphenoid, also unpaired. The major paired bones forming the side of the braincase are the prootic ("in front of the ear") caudally and the laterosphenoid rostially. These bones enclose a series of openings for cranial nerves and vessels. They contact the underside of the skull roof dorsally and ensure that the brain is well protected.24

The paired bones of the palate are thin, complex, hard to understand, and rarely seen because they are inside the skull. We shall content ourselves with knowing they exist and leave it at that. They are the pterygoids and the palatines. The long, unpaired vomer is rarely identified, but our Chasmosaurus has a nice one.

After all this brain work it is nice to switch to the lower jaws, which are pretty straightforward. Come to think of it, straight forward is the direction they run, from the jaw joint at the back to the beak in front. Each jaw is quite separate from the other, and each is articulated (joined) with the rather heavy predentary bone in a joint that may have allowed some degree of independent movement. The principal bone of the lower jaw is the dentary, the tooth-bearing bone.25 In Chasmosaurus the dentary is 42 cm long, and at the back it has a heavy, erect process, the coronoid process, that stands 16 cm above the lower border of the jaw. The coronoid process is a lever on which the jaw-closing muscles insert. The body of the dentary is rounded laterally, whereas the inner surface has a series of vertical grooves that represent the positions for each of the vertical columns of teeth. There are twenty-eight tooth positions in both the upper and lower jaws. We will have more to say about teeth in the last chapter. The other bones of the jaw are much smaller and are joined rather loosely; if an isolated lower jaw is found as a fossil, the smaller bones are usually missing because they have fallen off. The small bones are a bit of a jigsaw puzzle. The splenial is a long, thin bone that covers an open groove on the lower inner side of the dentary. The articular is a wide, cup-shaped bone that forms the joint with the skull.26 The cup is a socket to receive the condyle of the quadrate. The articular is very important, because it forms the hinge by which the jaw rotates open. The articular is supported underneath by the angular, and the prominent bone that links the articular and angular to the dentary is the surangular ("on top of the angular"), which is a tall bone that forms the rear portion of the coronoid process.

There—we have done it. The skull we have before us measures an impressive 147 cm in length. This size is routine, not exceptional, for horned dinosaurs, which had very large skulls. Of course the "true" skull, that is, the part common to all dinosaurs, was only half that length. The great length of these skulls is accounted for by the frill, the advertising structure. When the skull is joined to the skeleton, the total length of the animal is 4.9 m, or 16 ft 2 in. We have in fact assembled an entire skeleton from a pile of 312 bones (Fig. 2.11). It was, I admit, hard work. Even though I have studied dinosaurs for many years, I have never done a job like this before. I admit I could not have done it without the help of some excellent books and technical papers. Newton

Dinosaur Skeleton
FIG. 2.11. Skeleton of Chasmosaurus. (Bruce J. Mohn/ Robert Walters.)

was not just kidding when he stated, with modesty that was completely out of character for him, that the reason he saw farther was that he stood on the shoulders of giants. Science really is an additive process; we benefit from the discoveries and errors of those who have gone before us. It is not fair to criticize the gaffes of scientists from an earlier era who did not have the benefit of all that we know today.

There tends to be a sense of wonder that paleontologists are able to reconstruct skeletons at all. Admittedly it does require a base of knowledge, but the thing to remember is what vertebrates have in common with each other rather than what differs. The overwhelming pattern of a common body plan among all vertebrates is part of the evidence that convinced Charles Darwin that evolution had occurred; that is to say, the wonderful diversity of living things has arisen by the modification over time of a single common ancestor. Most of the bones of the dinosaur skeleton are found in our own bodies as well! If we can recognize a bone as a humerus, then we know, whatever animal it comes from, that it articulates with the shoulder joint at one end and at the elbow at the other. A skilled paleontologist can usually determine what part of the body a bone comes from, even if he or she has never seen that kind of animal before. We are not geniuses (at least I am not)—we just use the clues before us.

So we have completed a rather detailed survey of anatomy. Are you still with me? Don't worry—you won't be tested! But we have learned a very important vocabulary that will make communication much easier. The effort is not wasted. We do not have to learn a separate vocabulary when we speak of Triceratops or Protoceratops, or of Tyrannosaurus, or of alligators, or even of mammals. The features held in common among all vertebrates are far more important than the differences. But as the French say, vive la differenceA.

Was this article helpful?

+1 0

Post a comment