Struthiomimus altus (Lambe, 1902)

Osborn, H. F., 1917, Skeletal Adaptations of Ornitholestes, Struthiomimus, Tyrannosaurus, Bulletin of the American Museum of Natural History 35, pp. 733-771 : 744-760

Description of Skeleton of Struthiomimus altus .

Text-figs. 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , Plates XXIV , XXVI .

Proportions. In comparing this skeleton with that of Ornitholestes ( Plate XXVI ) we are immediately impressed with the extremely small size of the skull, elongation of the neck, abbreviation of the back, relative elongation of the tibia and of the tarso-metatarsus. These are the struthious analogies.

Non-struthious is the balancing power of the tail. The vertebrae behind the acetabulum exceed those in front of it by nearly a meter, and the posterior half of the tail is very stiff and rod-like and was thus doubtless used as a balancing organ in rapid cursorial motion.

As compared with Ornitholestes the fore limb is of the same relative length, but the manus is a very strong and fully functional organ.

Skull. The unique feature of the skull is the total absence of teeth and indications that the premaxillaries and dentaries were sheathed in narrow horny beaks somewhat similar to those of Struthio . The skull as a whole ( Fig. 5A, B ) is only one-third longer than that of Struthio although it is attached to a vertebral column over thirteen feet in length. The dentary is relatively deeper and more powerful than that of Struthio and apparently closes on the inner side of the premaxillaries. All the areas of muscular attachment are very slight as compared with those of the carnivorous dinosaurs. The muscles of the jaws are relatively feeble as compared with those of the carnivorous dinosaurs but more powerful than in Struthio .

The relations of the dentary, surangular, angular, articular, lachrymal, jugal, quadrate are totally different from those of Struthio and indicate direct derivation from the theropod dinosaur type. The chief distinctions from the cranium of Allosaurus and Tyrannosaurus ( Osborn, 1912, p. 28, figs. 25, 26 1) are: (1) the great size of the orbit, which is relatively as large as in Struthio ; (2) the constriction of the post-temporal fossa; (3) the vertical elongation of the quadrate-quadrato-jugal; and (4) the anterior inclination of the lower end of the quadrate as distinguished from the backward inclination of the same element in Allosaurus . The cranium thus presents a very wide contrast to that of Allosaurus and the widest possible contrast to that of Tyrannosaurus , indicating a profound change of function.

Cervicals. The cervical centra are uniformly elongate and the cervical ribs, instead of being loosely attached as in Ornitholestes and in the typical Theropoda, are abbreviated and closely coalesced to the sides of the centrum and neural arch in C3-10. Cervicals 6-11 are amphicoelous. C12 is procoelous. C4 is slightly procoelous. C3 is opisthocoelous. This is an indication of a long, flexible, and powerful neck, as in Struthio .

Thoracics. The thoracic or dorsal vertebrae, T1-13, are distinguished by elongate bicipital ribs, the anterior pair of which (T 1-T 2) are freely movable, while in T 3-T 10 the thoracic ribs are connected below with the sternal rib elements. In T 12 and T 13 the ribs are not observable in this specimen, but were probably attached to the inner side of the anterior crest of the ilium. The centra of T 1-3 are preserved and seem to be gently opisthocoelous; from T4-13 the centra are not observable; the neural spines indicate progressive increase in size and abbreviation in length.

Sacrals. The sacral centra cannot be clearly distinguished; the indications are that there were five sacrals closely coalesced. S 3-?5 are visible behind the acetabulum (compare Tyrannosaurus, Plate XXVII ). The supposed S 5 is coalesced laterally with the postero-internal plate of the ilium. Superiorly the sacral spines unite in a continuous single, closely compressed plate, wedged in between the iliac borders (compare Tyrannosaurus , Fig. 17 ).

Caudals. Seventeen caudals are preserved in continuous series in this specimen, Amer. Mus. 5339 , the typical theropod characters of which are clearly shown in Plate XXVI . Then follows an unknown interval in which five vertebrae have been inserted, behind which fifteen vertebrae are added from the specimens American Mus. Nos. 5355 , 5262 . These vertebrae, C 12-C 26, C 30, exhibit the remarkable rod-like elongation of the prezygapophyses, so that the apex overlaps the base of the prezygapophysis in front. The postzygapophyses are greatly reduced. The chevrons are elongate and rod-like also.

Sternal ribs. These elements consist of at least thirteen rows (corresponding with the thirteen thoracic ribs) each including a triple series on each side as represented in Fig. 6 and Plate XXVI , namely: (1) a short, very slender proximal series; (2) a slender median series about twice the length of the proximal; (3) a stouter ventral series, of about three times the length of the proximal, which terminate in club-shaped expansions in the median line. The sternal rib apparatus as a whole is exceedingly delicate and indicates a feeble development of the pectoral muscles.

Fore limb. Figs. 3 A , 8 . The scapula is slender and firmly coalesced with the coracoid. The humerus ( Fig. 7 ) is relatively slender as observed in two specimens ( Amer. Mus. Nos. 5339 , 5201 ). The small, backwardly directed head ( Fig. 7 C ) is supported by the lesser and greater tuberosities, which are somewhat more strongly developed than in the arboreal two-toed sloth, Choloepus didactylus ( Fig. 7D ). The shaft is relatively longer and more slender than in Choloepus and the distal extremity is relatively narrower and has less provision for muscular attachment than in Choloepus . The absence of muscular attachments on the pronator and supinator condyles of the humerus as well as the generally slender structure of this bone forbid the assumption of fossorial function for this peculiar limb. The ulna and radius are attached somewhat obliquely to the perpendicular axis of the shaft. A small sesamoid or vestigial Mtc. IV appears at the junction of the ulna and Mtc. III. The carpal series is represented by five separate elements, Fig. 3 A ; it includes the radiale, intermedium’, ulnare, carpale 1 and carpale 2, much flattened. Mtc. I—III are short and equal in length, and Mtc. II—III are firmly fitted to each other; Mtc. I and Mtc. II exhibit an incipient coalescence; they are synmetacarpal but not syndactyl.

The phalanges of Digit I, the pollex, are separate and divergent from those of Digits II, III which tend to be appressed and syndactylous; there are indications of more or less rotation and opposition or grasping power in the pollex.

Manus. The very unequal elongation of the phalanges is a peculiar feature of the manus; especially the first phalanges of Mtc. I, II, III are of totally different lengths. This is compensated for by the elongation of phalanx D II.2 which is unusually long. These three digits, therefore, through the abbreviation of D III and the elongation of D I are approximately isotridactyl as in Bradypus tridactylus ( Fig. 9 ). This arrangement is quite different from that of Ornitholestes hermanni ( Fig. 3 ) in which D I is reduced but D II is very much more elongate than D III. The lateral motion of these phalanges is partly provided for in the joints between the metacarpals and the proximal phalanges. The remaining interphalangeal joints tend to confine the digits exclusively to an opening and closing movement as in many edentates or strict flexion and extension such as is also observed in Ornitholestes and in Bradypus . The distal phalanges are extremely long, slender, and slightly recurved, with strong attachments for the flexor perforans, less recurved, claw-like, or raptorial than those of Ornitholestes ( Fig. 2 ). They are relatively longer and more slender than those of the greater or lesser anteaters and entirely lack the fossorial powers of the manus of these Myrmecophagidae.

Hind limb. The hind limb ( Fig. 8 ) is relatively robust, the femur ( Fig. 8 F ) having on a small scale the general form and areas of muscular attachment characteristic of the larger theropoda such as Tyrannosaurus ( Fig. 18A 1), with greater relative elongation. The tibio-tarsus ( Fig. 10 T ) is also a powerful bone, relatively more elongate than the femur. The fibula ( Plate XXIV ) is relatively slender.

The tarso-metatarsus, as previously described in Ornithomimus ( Figs. 10 , 11 ), consists of three closely coalesced median digits, D II-IV, in which Mts. III is thrust backward at the proximal end so that it only appears on the posterior face, while it is thrust forward at the distal end so that it makes up the greater part of the anterior face of the pes, in which D III is much the longest digit. Metatarsal V is greatly reduced, extending one-third down the shaft of Mts. IV.

The terminal phalanges are very slightly recurved, less elongate and pointed than in Dinornis among the struthious birds and slightly more pointed than in Struthio . This structure points to strictly cursorial rather than to raptorial function in the pes.

Habits of Struthiomimus .

In view of the ostrich-like structure of the skull and of the pes, and the partly suspensory, partly grasping structure of the fore limb and manus of Struthiomimus , it is very difficult to form a consistent hypothesis of the habits of this remarkable animal. It may be compared with certain of the lizards, the struthious birds, the tree sloths ( Bradypus, Choloepus ), and the Aye-Aye ( Cheiromys).

Comparison with lizards. 1 The habit of lizards of running upon the hind legs ( Figs. 10 , 11 ) with the body reared upward has been observed among representatives of three families, namely, the Old World Agamidae (Agamas) , the New World Iguanidae (Iguanas) and Teiidae (Tegus, Race Runners). (1) The two first named families have a distinct affinity, arising either from parallelism of development among varied genera or from actual relationship as alleged by some writers. The systematic separation of the Agamidae and the Iguanidae is mainly by the dentition, which is acrodont in the former and pleurodont in the latter. The family Teiidae is of remote relationship to both of the first named families. (2) Among the species of lizards observed at times to rear and use the hind legs only for locomotion are highly active types, large and small, that live under various conditions. The habit has been observed among large arboreal species as well as among lizards of diminutive size, several of the latter frequenting the deserts. (3) The lizard never rears, in fact, appears unable to do so, unless moving at a high rate of speed. The bipedal trait may be indicated immediately after the beginning of a dash over the ground, or the lizard may rear upward after it has run a considerable distance. In all cases the animal drops to all fours as its speed materially decreases. During the reared running pose the front legs appear to involuntarily droop forward and not to be used as balancers, but the position of the tail indicates the marked importance of this organ in balancing and a voluntary rearing or curving upward of the tail in order to keep the body upright. Mr. Ditmars has not observed the habit except in long-tailed species. (4) Lizards resorting to anthills are mostly burrowing and more or less limbless forms which present no analogy with Struthiomimus . (5) The grasping powers of the manus are developed in the tree-living chameleon, the fingers terminating in strong claws, two on each grasping knob, which should be compared with those of Struthiomimus ..

Running gait of Struthiomimus and Ornithomimus . From these studies of the lizards the running gait of Struthiomimus is suggested in the pose represented in Fig. 14A in which the rod-like tail balances the anterior half of the body. The estimated total length of the vertebral column being M 4.085, the length of the tail is estimated at M 2.525; thus the postacetabular vertebrae are slightly longer than the pre-acetabular. The hind limbs are relatively longer than in any of the running lizards but relatively shorter than in the ostriches. It is noteworthy as compared with the running lizards that the ratio between the length of the fore and hind limbs is less in Struthiomimus . Its very awkward position in the supposed quadrupedal gait is shown in Fig. 14B .

Insectivorous theory of C. William Beebe. The ornithologist, C. William Beebe, has suggested that the feeble fore limbs of Struthiomimus may have been adequate for attacking sandy and gravelly anthills, as represented in Fig. 14B . There are three objections to this theory: first, the fore limb gives no evidence of any fossorial musculature; second, the beak affords no evidence of a protrusible tongue analogous to that of the Myrmecophagidae ; third, the terminal phalanges of the hind limb afford little evidence of power to scratch or dig but are rather of the cursorial type.

Sea-shore theory of Barnum Brown. Influenced by the fact noted above that the remains of Struthiomimus and Ornithomimus are invariably found in littoral deposits, Barnum Brown has suggested the theory that this animal waded along the shores, feeding upon small crustaceans and molluscs, using its long anterior claws partly to scrape the sand and partly to seize these shore-living animals. The objections to this theory are three. First, neither the beak nor the terminal phalanges are adapted to seizing an actively moving crustacean prey of any kind although they may have been adapted to securing sessile Mollusca. Second, the feeble musculature of the forearm is non-fossorial, and non-raptorial; that is, it does not indicate powers of pronation and supination of the manus. Third, as pointed out by Beebe, neither the structure of the beak nor of the phalanges of the pes is analogous to that of the shore-living birds (grallatores) which partly live upon small invertebrates. Brown points out that while the attachments of the muscles are but slightly developed and the arm is of feeble power as a whole, the manus shows wonderful adaptability of finger movement, since while Mtc. II and Mtc. III are coalesced D I has an articulating face with Mtc. I which allows of a rotating movement and the opposition of DI to DII, III. The points noted are to Brown conclusive proof that this dinosaur in securing food depended chiefly on the movements of its neck and the prehensile power of the digits of the manus and that crustaceans were the favorite article of food, which the construction of the neck and manus were well adapted to capture.

Ostrich-like theory of Struthiomimus, Osborn . This theory, which seems the most probable one to the author, is depicted in Fig. 15C , namely, an herbivorous browsing pose in which the ostrich-like beak, sheathed in horn, is represented as securing leaves and buds while the fore limbs, two digits of which are analogous to those of Bradypus tridactylus , are used in supporting the fore part of the body and drawing down the smaller branches; the grasping power of the thumb, D I, may have facilitated this action.

First, in defence of this theory, a chiefly suspensory and limb-grasping function of the fore limb is indicated by its analogy to that of the tree sloth, Bradypus tridactylus ( Fig. 9 ), the analogy consisting in the slenderness of the humerus and of the ulno-radius, the apparent slenderness and suspensory character of the muscles and especially the isodactyly of the manus, a highly distinctive feature of the tree sloths. Second, the birdlike and especially ostrich-like structure of the neck, which is totally unlike that of the carnivorous dinosaurs and very similar to that of the struthious birds especially, in which very rapid fore-and-aft motions of the head are observed during the act of browsing. Third, the bony structure of the upper and lower jaws indicates the possession of a horny beak similar in shape to that of the browsing birds especially of the struthious type. There is no curvature of the beak suggesting analogy to the carnivorous birds, nor broadening of the beak suggestive of that of the wading birds (grallatores). Fourth, it appears probable that the pes was purely cursorial, adapted to dry or hard earth, since its analogy to that of the various members of the ratite order, such as Struthio , Rhea , Casuarius , Dinornis, is very close. Such feet are used primarily in locomotion, secondarily for the scratching of sand for the burying of eggs, and thirdly as weapons of defense; they are not adapted for assisting the fore limb and the beak in the prehension of food.

Notes on the Habits of Struthiomimus by Dr. William K. Gregory. (1) “Were the forearms used in pulling food toward the mouth, either up from the ground or down from the trees? Probably yes. (2) Were the forearms used in climbing up low trees? Probably no. (3) Were the forearms provided with a small patagium, which could be used in swift flight on the ground? [Hypothetical.] ”

“ The long slender forearm of Struthiomimus had a long reach and great freedom of motion at the shoulder, as shown by the form of the head of the humerus. The elbows were undoubtedly everted, in reptilian fashion. While the radius could not be twisted around the ulna so that true supination was very limited, yet merely by moving the elbow in toward the body the palms of the opposite manus were turned toward each other (partly supinated), while by moving the elbows outward from the body the manus was brought into the position of pronation and the opposite manus brought together. The wrist joint was very stiff, capable of very little extension; the three metacarpals were closely appressed; there was considerable motion between the metacarpus and the proximal phalanges. In the third digit the second and third metacarpals were closely appressed and in the way of coalescence, but the distal phalanges of the second and third digits could be sharply flexed. By virtue of the twisting of the first metacarpal the phalanges of the first digit could be turned partly toward the other digits so that the manus was to a certain extent prehensile. The ungual phalanges were large and falcate, deeply grooved at the sides and evidently bore stout and sharp claws, larger than those of a grizzly bear. The deep pits at the sides of the distal ends of the penultimate phalanges testify to the great strength of the ligaments connecting the distal and penultimate phalanges, while the large size of the claw phalanges and their swollen proximo-inferior bases indicate a powerful development of the flexor profundus (perforans). The manus as a whole, as observed by Professor Osborn, partly suggests that of the tree sloths, not only in its general form, but in the tendency toward coalescence of two of the successive phalanges, and in the large size of the claws. It differs from the sloth type, however, in the lateral mobility of the first digit and of the proximal phalanges of all the digits.”

“ The reconstruction of Struthiomimus errs in representing the forearm in the position in which it is mounted, namely with the elbow drawn well in toward the body and with the dorsum of the manus turned fully outward. That this position is erroneous is apparent from the fact that the head of the humerus is thereby wrenched out of the glenoid and appears exposed in the lateral view. By placing the head of the humerus fully within the glenoid, everting the elbows and opening the radio-humeral angle, the manus will be partly pronated.”

“We speak of the fore limb of Struthiomimus as ‘slender’, but it is so only because the longitudinal dimensions of the principal elements greatly exceed their transverse dimensions. The fore limb is, however, by no means weak. It is at least as strong as the fore limb of a man and was armed with very large sharp claws.”

“ With the elbows everted in reptilian fashion and the body swung downward from the acetabulum, the powerful manus could be brought near to the ground, although I think there was not sufficient mobility for the fingers to be extended so that the palm could rest upon the ground. But I believe that the animal could stoop down, and with its powerful claws tear up shrubs from the ground and perhaps scrape away the surrounding earth by the sweeping movements of the whole forearm after the fashion of a kangaroo. These heavy claws might also have been useful in tearing off the husks of fruits or for tearing and pulling down the branches and fruits of palms, cycads, figs, bananas, etc. Perhaps the claws served also as weapons of defense, if they were swiftly drawn across the flanks of an opponent. The pulling up of vegetation from the ground and the tearing of husks would be assisted by the broad, ostrich-like beak and the powerful neck, which was capable of strong vertical movements.” ‘

“ Some features of the fore-limbs might indicate that Struthiomimus was a climbing animal, but the large size of the body and the nonprehensile character of the pes are against this view.”

“ The jaws of Struthiomimus unquestionably suggest those of an ostrich, and the animal seems too large to thrive exclusively upon an insectivorous diet. Struthiomimus was not necessarily exclusively a vegetarian. The ostrich itself is said to be omnivorous in the wild state, feeding on lizards, small mammals, fruits, berries, seed grass, etc., and the slightly more powerful beak of Struthiomimus might have been equally adapted for a varied diet, including some animal food.”

“The fact that the remains of Struthiomimus were found in littoral deposits does not in my judgment indicate that the animal lived after the fashion of wading and shore-living birds.”

“In conclusion we can be sure only that Struthiomimus was a swift running biped; that in seeking its food it made well-coordinated use of its ostrich-like bill, powerful neck, and strong, heavily clawed manus; that a good deal of pulling and clawing was accomplished by these wide-reaching fore limbs.”

“ The animal seems to be too big for arboreal life, and its hind limbs were built rather for great speed on the ground, purely as a protective function. It seems possible that the fore limbs bore a narrow patagium, which might be of use in rapid flight along the ground as indicated in the accompanying restoration ( Fig. 16 ). By throwing the head and forearms well forward the thrust of the femora would be turned more forward than upward, while at the same time the forearms if provided with patagia would serve to keep the fore part of the body from pitching downward.” 1

Author's conclusion. The above observations by the several expert anatomists who have cooperated with the author in attempting to solve the very difficult problem of the habits of Struthiomimus indicate a very great diversity of opinion and interpretation. The author adheres to the opinion, very gradually formed, that Struthiomimus presents prevailing analogies of structure in the head, neck, and feet, and consequently of habit, to the ostriches. Its fore limb structure, entirely unique among reptiles, is analogous partly to the suspensory type of fore limb such as we observe in the tree sloths, with the exception of the phalanges of the pollex, Digit I, which are set apart from the other digits and exhibit a grasping or opposable function.

Comparison of the Struthiomimus manus with that of the fossorial anteaters ( Myrmecophagidae ) shows that the structure is totally different, for anteaters and, in fact, many other fossorial animals are anisodactylous, that is, the different digits vary greatly in length, in size, and in the development of the claws, whereas in Struthiomimus we have the isodactyly peculiarly suited to the suspensory, limb-grasping power of arboreal types. This does not preclude fruit- and shrub-grasping powers, but appears to exclude the grasping of living or actively moving prey.

Figures

Fig. 2. Incomplete manus of Ornitholestes hermanni; Digit I is missing in this specimen. Amer. Mus. No. 587. Palmar (A) and internal (B) views, one-half natural size. After Osborn, 1903.

Fig. 3. Composition of complete manus (B) palmar view of Ornitholestes hermanni. Amer. Mus. 587, partly restored (Digit I) from Amer. Mus. 619, type. One half natural size. Manus of Struthiomimus altus, Amer. Mus. 5339. (A), dorsal aspect; A', palmar aspect, r, radiale, i, intermedium (?) coalesced with cc, centrale. Cl, C2, p. C3 carpalia and (?) p. tendinous ossicle. One third natural size.

Fig. 4. First restoration of the skeleton of Struthiomimus altus, Amer. Mus. 5339. Digit I is set too close to digits II, III. One-thirtieth natural size.

Fig. 5. Skulls of Struthiomimus and Struthio compared. A Portions of the skull and jaws of Struthiomimus altus, Amer. Mus. 5339, showing the left dentary, also the outer and inner aspect of the right dentary. B Skull of Struthio. Both figures one-half natural size

Fig. 6. Detail of the actual arrangement of the abdominal ribs in Struthiomimus altus, Amer. Mus. 5339, as preserved on the left side of the thorax. One-fifth natural size.

Fig. 7. Humerus of Struthiomimus altus and of Choloepus didactylus compared. A Crushed left humerus of S. altus, Amer. Mus. 5339. B Normal left humerus of S. altus, Amer. Mus. 5201. C Right humerus, anterior and posterior views of S. altus, Amer. Mus. 5201. D Right humerus of two-toed sloth, Choloepus didactylus, Amer. Mus. 35513, posterior view. All figures two-ninths natural size.

Fig. 8. Fore and Hind limb bones, H, U, R, of Struihiomimus altus, Amer. Mus. 5339, to same scale. Palmar aspect of left fore limb. F Left femur, internal aspect. T Left tibio-tarsus, internal aspect. All figures two-ninths natural size. Manus more correctly drawn in Fig. 3.

Fig. 9. Skeleton of the three-toed Sloth, Bradypus tridactylus. After De Blainville.

Fig. 10. Manus of two-toed Sloth, Choloepus didactylus. Palmar aspect. Natural size.

Fig. 11. Left pes of Struthiomimus alias, Amer. Mus. 5339. Internal, posterior, anterior, external aspects. Mts. II, III, IV, with three functional digits, Mts. V reduced. Two-ninths natural size. Compare Fig. 3a.

Fig. 14. Restorations of Struthiomimus altus, Amer. Mus. 5339. A Running pose suggested by Osborn. B Insectivorous theory, anthill-attacking pose suggested by C. William Beebe. One- fortieth natural size.

Fig. 15. Restorations of Struthiomimus altus. C Herbivorous, tree-browsing pose, suggested by Osborn. D Carnivorous, shore-living pose, hunting for Crustacea and small Mollusca, suggested by Barnum Brown. One-fortieth natural size.

Fig. 16. Running pose of Struthiomimus, with a rudimentary patagium as suggested by W. K. Gregory. One-fortieth natural size.

Fig. 17. Tyrannosaurus and Struthiomimus mounted skeletons, reproduced to same scale, about one-seventy-fifth natural size, and showing relative size and proportions of the fore and hind limbs.

Fig. 18. Mounted skeleton of Tyrannosaurus rex Osborn, anterior and lateral views. One-sixtieth natural size. The structure of the manus is conjectural. The other parts are mounted by combination of the skeletons Amer. Mus. Nos. 5027, 973. The sternal ribs are omitted. The manus is represented as tridactyl, but it may be didactyl.

Plate XXIV. Anterior, lateral, and posterior aspects of the mounted skeleton of Tyrannosaurus rex, chiefly from Amer. Mus. 5027, partly from the type specimen Amer. Mus. 973. The sternal ribs are not restored. The structure of the manus is unknown as yet; the restoration of the digits is conjectural. About 1/53 natural size.

Plate XXVI. Reconstruction of the skeletons of Struthiomimus altus (left) and Ornitholestes hermanni (right). Struthiomimus 1/10, Ornitholestes 1/6 natural size, The Ornitholestes restoration replaces the original restoration by Osborn in 1903 which is very faulty. The Struthiomimus, Amer. Mus. 5339, mount has the distal end of the tail restored from Amer. Mus. 5355; dotted vertebra from Amer. Mus. 5262, In both restorations the pollex is too closely appressed to the other digits, see Fig. 3.

Plate XXVII. Pelvic arcade and vertebral structure of the second chief specimen of Tyrannosaurus rex, Amer. Mus. 5027, discovered in 1908. An orthogonal projection executed on a very large scale and reproduced one-twelfth natural size. C 1-C 10 cervical series, D 1-D 13 dorsal or thoracic series, S 1-S 5 sacral series, Cd 1- Cd 53 caudal series. The caudals actually preserved are shaded; those drawn in outline are conjectural and restored. The total number of caudals is conjectural

Fig. 3a. (A1-A3). Genotype specimen of Ornithomimus velox Marsh. Yale Mus. 542. (A1), dorsal aspect of the type left pes, superior aspect of Metatarsals II-IV. (A2) posterior aspect (upper) and exterior aspect (lower) of metatarsals II-IV. (A3) Type tibio-tarsus, anterior, inferior, posterior aspects. (B) Portion of right manus of another individual (Yale. Mus. 548) associated by Marsh with Ornithomimus velox; (upper) metacarpals I-III as they appear in the matrix, (lower) as placed in natural position. (C) Left pes of Struthiomimus altus, genotype specimen, Amer. Mus. 5339. Anterior and exterior aspects of digits II-V. One fourth natural size. After original drawings by the present author. One-fourth natural size.