Describer

Hou, Zhou, Gu and Zhang, 1995

Time

Jurassic Late Cretaceous Early Barremian

Classification

Saurischia Theropoda Tetanurae Coelurosauria Maniraptora Avialae Confuciusornithidae

Fossilsite

Chaomidianzi Formation ( previously referred to as the lower section of the Yixian Formation), Sihetun Village, near the town of Shangyuanzhen, Beipiao (Beipiaoshih) Municipality Liaoning, China

Info

Confuciusornis sanctus ~/~ (Hou, Zhou, Gu and Zhang, 1995) > Confuciusornis chuonzhous ~/~ (Hou, 1997) > Confuciusornis sunia ~/~ (Hou, 1997) > Confuciusornis dui ~/~ (Hou, Martin, Zhou, Feduccia and Zhang, 1999) > Confuciusornis meidus ~/~ [Nomen Nudem] (Hou, 1997) > Confuciusornis shuzhi ~/~ [Nomen Nudem] (Hou, 1997)

Hundreds of complete skeletons

Hou L.-H.; Zhou Z.; Martin, L.D.; and Feduccia, A.(1995). A beaked bird from the Jurassic of China. Nature 377: 616-618.

Confuciusornis sanctus provides the earliest evidence for a beaked, edentulous bird. The associated wing skeleton retains the primitive pattern found in Archaeopteryx, including a manus with unfused carpal elements and long digits. Two leg skeletons from the same site also show an Archaeopteryx level of morphology, and provide the earliest indisputable evidence for a covering of body contour feathers. These specimens provide evidence for either an undiscovered pre-Archaeopteryx or a rapid, post-Archaeopteryx evolution in birds. As the first Jurassic birds to be described from outside Germany, they show that birds with long fingers terminating in large recurved claws were widely distributed. They are not found in the Early Cretaceous sediments of the same region, where there is a diverse assemblage of more advanced flying birds with smaller fingers and claws. The postcran-ial structure of Archaeopteryx and Confuciusornis seems to be adapted for climbing tree trunks and may have disappeared near the end of the Jurassic.

Confuciusornis sanctus Hou, Zhou, Gu, and Zhang, 1995 Specimens: In the initial publications (Hou et al. 1995A, 1995B, Hou et al. 1995) three Jurassic specimens with complete feather impressions were erected as the species (IVPP specimen numbers V10918, V10919, and V10895). However, subsequent observations of the incomplete hindlimb of specimen 10919 indicated rather significant distinctions from the type. Most recently a relatively complete Confuciusornis skeleton has been recovered that displays a hindlimb resembling that on V10895 and confirms the distinct discrepancies of the V10919 specimen. Therefore, this specimen will hereby be recognized as a distinct species with a description provided later in the text. Stratigraphic position and age: Interbedded calcareous shales of the Late Jurassic basal Yixian Fm. Locality: Jianshangou by the village of Shangyuanzhen, Beipiao Municipality (Beipiaoshi), Liaoning Province. Ammended diagnosis: The size of the Eichstaff Archaeopteryx specimen. Edentulous but with regularly patterned grooves in the jaws indicating the presence of a keratinous beak. Orbit is large and antorbital fenestra is circular. Proximal end of humerus is expanded with a pneumatoceoel present. Ungual phalanx of the first digit is long and robust, manus phalanges are slender and unreduces, and carpals are unfused with metacarpals. Ischium is robust with a slightly expanded sital end and a dorsal process near its proximal end. A fifth metatarsal is present. Pes talons are large and recurved. Summary description: Morphological studies of Archaeopteryx have been conducted for over 100 years and in the past 20 years numerous monographs and other works have been published upon it. Obviously, due to deficiencies in preservation, the interpretations of several morphological characters are inconsistent. Due to the absence of a complete squamosal, the precise configuration of the jaw articulation is in doubt. An additional ongoing debate concerns the nature of the pectoral girdle and the effective strength for flight. Confuciusornis postdates Archaeopteryx but due to its autapomorphic characters and the characters it shares with Archaeopteryx, is a new member of the Sauriurae represented by abundant specimens of Late Jurassic primitive birds which provide data for the comprehension of Jurassic avian evolution and diversity. They also provide insights into the origin of Aves. In this study, documentation of the morphology of Archaeopteryx was based upon two casts provided to IVPP by the British Museum of Natural History and the Berlin Museum. In addition, reference is made to a restoration of the skeleton by Martin (1991). Additional documentation was derived from the paleornithological literature.

1. Occipital region. Several occipital elements enclose the foramen magnum and form the posterior cranial wall. The foramen magnum on both Confuciusornis and Archaeopteryx are identical, being situated posteriorly and resembling the configuration of their reptilian ancestors. Only recently has the occipital region of Confuciusornis been prepared sufficiently for description. The left side is poorly preserved but the right side including the foramen magnum is well represented. Dorsal to the foramen magnum is a relatively substantial supraoccipital which, in posterior perspective, has a trapezoidal posterior wall. On its dorsal section a posteriorly extended crest surpasses the occipital plane and becomes confluent with the dorsal supraoccipital-parietal crest to represent the highest point on the cranium. Exoccipitals form the lateral walls of the foramen magnum, are larger than the supraoccipital, are transversely expanded, and constitute the largest pair of elements on the posterior wall of the cranium. Their most prominent characteristic is the formation of an exoccipital tympanic bulla, which is homologous to the exoccipital tympanic wing on extant avian taxa. At the lateral side of the tympanic bulla is a large and broad paroccipital process. Upon the anterodorsal margin of this process is the foramen for the hypoglossal cranial nerve (XII). Ventral to the foramen magnum is a relatively small basioccipital, although its morphology and that of the occipital condyle are obscured due to compressional distortion.

2. Auditory region. The auditory region is composed of three tympanic elements which bear an intimate relationship to the elements on the occipital region. The epiotic is a long, linear, and oblique element located on the lateral supraoccipital and dorsomedial margin of the exoccipital and contacts the ventral parietal. The opisthotic is posterolateral to the epiotic and is isolated from it posteriorly by the exoccipital process. It is a lamelliform element that is rather similar to that described for Archaeopteryx by Martin (1991). The prootic is difficult to distinguish due to being subjected to compressional distortion, but it should lie anterior to the epiotic and medial to the squamosal with the exoccipital lying posteroventral to it.

In the general design of the vertebrate skull, the squamosal is considered a portion of the zygomatic region, but during the course of avian evolution it has become a pair of elements that have undergone extensive modification, and are located on the dorsal angle of the cranium close to the auditory region. On Confuciusornis each is well developed and lies posteroventral to the frontal and ventrolateral to the parietal to compose the posterior wall of the neurocranium; it represents a principal component of the temporal fossa. In morphology, it is a trapezoid plate with a ventral margin that contributes to the articular surface in combination with the quadrate. This is in distinct contrast to that of Archaeopteryx.

3. Basal plate. The basal plate of the avian neurocranium is composed of the basisphenoid with its associated processes and the rostral process of the parasphenoid. However, the most posteroventral wall of the neurocranium on true birds and on Confuciusornis consists of the horizontal plate of the basioccipital. The Confuciusornis basisphenoid and basioccipital have been somewhat displaced in position. The basisphenoid is relatively short, and the pair of anterior basipterygoid processes are extremely small. Anterior to the basisphenoid, the rostral process of the parasphenoid is well developed as being relatively long. Absent, however, is the typical pituitary fossa (sella turcica) between the basisphenoid and the rostral process of the parasphenoid, but an interorbital septum is extremely well developed.

4. Mid-parietal region. The dorsal midportion of the neurocranium, or the parietal region of the skull, is composed of the parietal and frontal. On Confuciusornis these elements resemble those of Archaeopteryx and are not fused. The parietals are the principal posterodorsal elements of the cranium and are represented by two lateral elements but are incompletely preserved due to compressional distortion. The anterior parietal is thinly lamelliform, forms the dorsal surface of the neurocranium, and contacts the frontal with a broad suture line. The posterior portion of the parietal forms the posterior crest of the skull and increases in thickness to form a V-shaped laterally extended crest in the occipital region. The lateral margin of this crest contacts the posterior margin of the squamosal.

The frontals of Confuciusornis are narrow and elongated anteriorly but broad posteriorly, which differs from Archaeopteryx but is rather similar to the frontals described for Protoavis from the Late Triassic of the US, which has yet to be acknowledged as a legitimate member of the class Aves. Only the base of the frontal is not preserved. On the right side it is relatively well preserved and is overlain by the parietal. A large portion of the lateral frontal constitutes the medial wall of the orbit. The anterior frontal process extends along the interorbital region to the ventral nasal and forms the thin and narrow posterodorsal wall of the nares. The length of the anterior process greatly exceeds that of the base and there is a conspicuous unfused boundary line between the two frontal processes.

Orbitosphenoid

This region is composed of the circumorbital and palatal series of elements. The circumorbital consists of the prefrontal and lacrimal, the former of which is represented by a pair of
irregular triradial elements (only the left is visible on the type) located lateral to the anterior process of the frontal and posterior to the lacrimal. Two of its branches extend to the lateral side of the frontal, are short, thick, and form the anteromedial wall of the orbit. The branch that extends anteroventrally, or toward the ventral lacrimal, is relatively large and elongated with a terminus that fuses to the main body of the posteroventral lacrimal and forms a portion of the anterior wall of the orbit.

The long, broad lacrimal forms the anterior wall of the orbit and is the single element that separates the antorbital fenestra from the orbit. At the long and broad anterior margin of this element a crest extends from the margin of the frontal anteroventrally and distinctly defines the anterior orbital outline. The posterior edge of this crest is sharp, thin, expansive, and lamelliform. The palatal complex consists of paired palatines, vomers, and pterygoids. The palatine has been compressionally distorted, obscuring its morphology, but it is located posterior to the palatal process of the maxilla. Expansively broadened at its base, it becomes a slender and elongated maxillary process. Its anterior margin contacts the palatal process of the maxilla with a tight suture line and posteriorly it is in contact with the pterygoid. The medial margins of the left and right basal portions extend anteromedioventrally to form a pair of short medial sections. The dorsal basal portions constitute the rostral process of the parasphenoid.

The vomer is the longest element at the midpoint of the anteroventral cranium. This narrow and long horizontal plate is posterior to the midline of the palatal process of the maxilla. Lateral to it lies the maxillary process of the palatine. Dorsal to the vomer is the olfactory canal. The internal nares should lie at its posterior terminus and medial margins of the palatine, although this cannot be documented due to the affects of compressional distortion. The pterygoid constitutes a major element of the medioventral neurocranium and tightly unifies the anterior and posterior portions of the skull. It is shaped as a curved rod with three contact surfaces or individual processes. The anterior process contacts the posteromedial margin of the palatine. The second process lies posterior to the anterior process and anteromedial margin of the pterygoid where it contacts the basipterygoid process. The third process ascends from the pterygoid to contact the quadrate. At the anterior branch of the quadrate there is a suture line (distinct on the left side) that very possibly represents a contact with the quadrate process of the pterygoid.

Ethmoid

A pair of nasals is visible at the anterodorsal portion of the cranium as a pair of elongated rectangular plates that are aligned along the midline of the rostrum and compose the majority of the
dorsal olfactory canal. Their anterior margin contacs the short nasal process of the maxilla, laterally they contact the expansive medial margin of the maxilla, posterolaterally they fuse to the prefrontal, and posteriorly they terminate at the anterodorsal frontal.

7. Visceral cranium. This region is composed of the mandibular arch and hyoid arch, but because of the preservational conditions of the specimens only a portion of the mandibular arch is recognizable.
The Confuciusornis quadrate has been affected by compressional distortion but still appears extremely robust and extremely irregular in morphology; it has several radiating branches, and crosses a basicranial articulation toward its contact with the neurocranium. Posterodorsally, it is in contact with the squamosal (or the opisthotic process) and anterodorsally it contacts the pterygoid. Dorsally, it is still only a single ascending branch with two surfaces and it is not discernible whether there are anterior and posterior branches here. The posterolateroventral section fuses with the terminus of the quadratojugal and jugal. The main body of the quadrate is expansive to facilitate the mandibular musculature; there are several attachment surfaces and thickened inflations.

On the London specimen of Archaeopteryx the right quadrate is relatively well preserved and resembles that of Confuciusornis, although its posterodorsal contact appears to be inconsistent. The quadratojugal of Confuciusornis resembles that of Archaeopteryx in being a relatively small triangular plate that is ventrolateral to the jugal. Its horizontal surface is relatively spacious and is fused dorsomedially to the jugal. Its vertical surface is relatively thin, slender, and extends posterodorsally; its distal end is configured as an acute and thin plate that fuses to the ventrolateral surface of the quadrate. The posterior quadratojugal and jugal form the osteroventral portion of the orbit.

The jugal is an elongated plate that forms the ventral margin of the orbit. Its anterior portion is relatively broad and becomes confluent with the maxilla. Just dorsal to the maxillary contact it is in contact with the ventral margin of the lacrimal. Posteriorly it attenuates to form an acute terminus that curves dorsally and fuses to the lateral quadrate. At its horizontal portion its ventrolateral side is in contact with the quadratojugal. In later species of Aves the postorbital process facilitates the musculature with the dorsal jugal.

The maxilla composes the principal region of the rostrum and consists of a pair of large elongated and expansive anterior cranial elements with a complex morphology. Its four components are the maxillary body, nasal process, jugal process, and palatal process. From dorsal perspective, the anterior maxillary body is expanded and the posterior portion is narrow. Anterolaterally there are two nutrient foramina. There are shallow striations approaching the midline which confirm the presence of a keratinous beak. The posteromedial margin extends posteriorly toward the dorsal midline to compose a narrow nasal process. Its posterolateral portion extends posterodorsolaterally to form the jugal process. The anterolateral side of the main body of the maxilla curves ventrally to form the lateral margin of the dental trough.

The ventral palatal process is visible as an expansive and flattened plate with its anterior margin in contact with the palatal process of the premaxilla, its anteromedial end in contact with the vomer, and its posterior portion in contact with the palatine. The premaxilla is irregular in shape and shaped much like an arrowhead. Its components are the main body, the dental trough process, and the nasal process. Its most characteristic feature is the systematically grooved ornamentation along the dorsal midline of the main body in addition to the presence of small lateral striations. This is further evidence that Confuciusornis possessed the first, although extremely short, bird beak. The anteroventral premaxilla is composed of a short dental trough process, but whether or not it bore dentition is unknown due to breakage (most recent specimens excavated confirm that the dentition is completely lost). The palatal process is broad and spacious with a posterior margin in contact with the anterior margin of the palatal process of the maxilla. The nasal process is the longest process of the element and forms an ascending branch which runs dorsally along the midline to form the separation of the external
nares. The terminus is inlaid with the nasals.

Cranial measurements of Confuciusornis sanctus (mm).
Length of skull 50.0
Breadth of skull 27.0
Orbital length 16.0
Orbital breadth 14.0
Antorbital fenestra length 9.0
Antorbital fenestra breadth 5.5
Antorbital fenestra to anterior rostrum 14.0
Premaxilla length 9.0

The postcranial skeleton of Confuciusornis is not as completely preserved as that of Archaeopteryx. Portions of the vertebral column are missing, the pectoral girdle is not preserved, and the forelimb is incomplete. Elements of the forelimb that are represented include the proximal humerus, which is extremely inflated and transversely expanded and has a head that is large and crescentic at its apex. The breadth of the proximal humerus exceeds half the humeral length, and has medial and lateral processes that are bluntly conical. At the midsection of the proximal end is a relatively large pneumatocoel which represents the oldest documentation of this character in the class. Its precise function is enigmatic although obviously it is related to facilitating flight and is a modification for weight reduction. The humeral shaft is relatively gracile and straight. In Archaeopteryx not only are there no adaptations for weight reduction, but the proximal humerus is narrow and long with an extremely small head. Clearly, these two are distinct because the latter genus lacks any complex derived morphological features for the adaptation for flight as in later avian genera.

Another significant character of the Confuciusornis forelimb is the particularly well developed ungual on digit I: its length exceeds that of the proximal phalanx, however its curvature is not as intense as in Archaeopteryx. The data and conclusions provided by Feduccia (1993) indicate that both Archaeopteryx and Confuciusornis had maintained manual and pedal ungual phalanges that were equivalent in their range of functional ability for perching and climbing. It is obvious on the Confuciusornis specimen block that there was a keratinous sheath covering the ungual that extended its length by approximately one-third.

Neither distal ulna nor radius bear multifaceted articular surfaces as noted on other genera of Aves and instead they bear condyles like ancestral reptiles for articulation with carpals. There are at least three relatively well preserved carpals. Three metacarpals are present: metacarpal I is short and broad. Metacarpal II is particularly robust and has a large articular facet for digit II. Metacarpal three is gracile and long, being the thinnest in the series, with a diameter less than half that of McII although its terminus is expanded. The first phalanx of digit I is the longest among all the individual phalanges. Digit II is equivalent in length to McII and is the largest among all digits. Digit three is the smallest and weakest among the digits but bears three phalanges, more than the other digits. Its first phalanx is particularly short and small.

Only a single relatively well preserved specimen of the pelvic girdle is represented (V10895). It lacks only its anterior portion and in general morphology resembles Archaeopteryx in comprising three unfused elements. The ilium is the largest of the elements; it has an expansively broad lateral surface that forms the majority of the acetabulum. Its dorsal margin is slightly crescentic with low pre- and postacetabular processes. In Archaeopteryx the dorsal margin is basically flat and straight, its preacetabular process is high and its postacetabular process is low and acute. The ischium is relatively robust with a smooth and glossy surface. Proximoventrally a low crest extends posteriorly to the posteroventral terminus. At the proximodorsal margin posterior to the acetabulum, a dorsal process with an apex that contacts the ventral margin of the postacetabular process of the ilium and forms an acetabular-ischial fenestra. There is a thick inflation at the terminus of the dorsal margin to facilitate contact of the abdominal musculature. This thick inflation is shorter than the terminus of the ventral angle. Consequently, the terminus of the ischium lies on a dorsoventral oblique plane with a concave surface ventral to the thick inflation. There are at least two discrepancies from the London specimen of Archaeopteryx, the first of which is the bifurcated terminus of its ischium. Secondly, the dorsal process on Confuciusornis is quite distinct from all specimens of the former. Furthermore, many Early Cretaceous specimens have been recovered, but none of them bear this feature, and as such this is a taxonomic character worthy of note.

The pubis is incompletely preserved on V10895 as the distal end is missing, but a complete and distinct impression of the element is preserved in the matrix. The proximal end is relatively broad and spacious. Proximolaterally there is a small pubic process, the anteromedial margin of which composes the anteroventral wall of the acetabulum. The posterior pubis is thin and posteriorly oblique. The distal end is rounded but not expanded, nor is it fused with its distal counterpart. This is quite distinct from Archaeopteryx which has fused distal pubes that are expanded to form a boot. Only a single left femur is preserved on V10895 that has been shifted somewhat in position. This is an extremely stout element, and although the head has suffered some damage, the rest of its proximal region and remaining element are well preserved. The proximal region is extremely flattened, broad, and spacious. The shaft is very slightly anteroposteriorly compressed and has a small amount of curvature but not to the degree as noted on Archaeopteryx. The two distal condyles are more well developed than in Archaeopteryx and project conspicuously anteriorly with a distinct intercondylar fossae.

On the hindlimb of V10895 there is an elliptical, flat, and smooth patella preserved between the tibia and femur. This is an extremely interesting phenomenon in avian evolution. This element acts as a buffer of equilibrium during the articulation of the femur and tibia and strengthens the flexion of the hindlimb, particularly during climbing and during active feeding, or a long period of flexion as in the process of brooding of young or roosting. Despite the absence of a noticeable posterodistal patellar groove, Confuciusornis may be compared to Archaeopteryx as having a relatively strong power of flexion due to the presence of a patella. A patella is not documented on any Early Cretaceous birds in China, which may be due to its being highly autapomorphic and thereupon lost.

The tibiotarsus is the longest and most robust skeletal element in Confuciusornis. Its termini are both slightly expanded, the shaft is straight, and laterally there are feather impressions upon it. As it is preserved in lateral perspective, the complete morphology of the condyles is obscured, but from lateral perspective the proximal articular surface appears to be slightly convex, unlike other archaic avian taxa in which there are two differentially sized concave articular facets. The proximal end is also slightly more expanded than the shaft. The ventral margins of the distal condyles are rounded, do not expand posteriorly, and are slightly anteriorly projected.

From an anterior perspective it is difficult to discern whether or nor the condyles are triangular in morphology, and thus these features can not be compared to Archaeopteryx. Only a single tarsal is preserved on V10895 and it is not entirely distinct. Archaeopteryx retains three reduced tarsals, which again indicates that Confuciusornis is more derived. The tarsometatarsus of Confuciusornis resembles that of Archaeopteryx in its length being only slightly longer than half the length of the tibiotarsus. Moreover, metatarsal V is retained and only the proximal tarsometatarsus is completely fused. The tarsometatarsus is relatively broad; metatarsal I is extremely short and slender and is only in contact with the distomedial side of MtII, which is also relatively slender. MtIII is the most robust and longest of the metatarsals, MtIV is slightly thinner than MtIII but more robust than MtII. MtII is the longest of the metatarsals.

Four digits are present on the foot which has a phalangeal formula of 1-2-3-4. Digit III is the longest with the third phalanx as a relatively long and hooked ungual that has a relatively well developed flexor tuberosity at its ventral base and flexor grooves laterally that run to the terminus of the ungual, indicating that Confuciusornis was well adapted for relatively strong claw flexion. Comparison and discussion: Both Archaeopteryx and Confuciusornis are assigned to the Sauriurae (Hou et al., 1995) and share numerous characters. Both genera are recognized as primitive terrestrial birds despite the fact that Archaeopteryx is preserved in marine lagoonal
deposits while Confuciusornis is derived from fluvio-lacustrine sediments. The two genera are also recognized as generally contemporaneous.

If one considers Archaeopteryx to possess numerous autapomorphic or derived characters, then Confuciusornis possesses even more. The seven specimens of Archaeopteryx represent the most primitive species of Aves and to date are regarded as the only data representing the origin of the class. The discovery of Confuciusornis destroys the notion that Jurassic birds are rare. Concurrent with the discovery of Confuciusornis is another report from Korea of a new discovery of Archaeopteryx from Jurassic sediments, which reportedly consists of a forelimb. Strict attention will be provided to the results of research upon this new material, particularly with regard to its relationships to Archaeopteryx. Both Confuciusornis and Archaeopteryx share plesiomorphic characters for Aves, including the presence of feathers, absence of a postorbital, pelvic girdle rather conspicuously posteriorly oblique (moderately), and a fused proximal tarsometatarsus (Cracraft, 1988). However, both still retain numerous reptilian plesiomorphies, including unfused cranial elements on mature individuals, foramen magnum positioned posteriorly on the occiput, retention of three

Postcranial measurements of Confuciusornis sanctus (mm).
Preserved length 51.0
Proximal breadth 28.0
Humerus Longitudinal diameter of pneumatic foramen 7.0
Breadth at midshaft 7.0
Distal breadth 8.5
Preserved length 42.0
Proximal breadth 6.2
Ulna Breadth at midshaft 4.0
Distal breadth 5.3
Preserved length 45.0
Proximal breadth 3.5
Radius Breadth at midshaft 3.0
Distal breadth 5.0
Length 9.0
McI Breadth 3.5
Length 22.5
McII Breadth 3.5
Length 19.0
Proximal breadth 2.0
McIII Breadth at midshaft 1.1
Distal breadth 3.0
Length 20.5
Manus digit I Proximal breadth 4.5
Length of ungual and sheath 20.0
Phalanx 1 length 21.0
Manus digit II Phalanx 1 breadth 6.0
Phalanx 2 preserved length 10.5
Phalanx 1 length 5.0
Manus digit III Phalanx 2 length 14.0
Phalanx 3 length (preserved) 13.0
Length 33.0
Proximal anteroposterior breadth 5.0
Femur Anteroposterior breadth at midshaft 3.5
Approximate distal breadth 3.3
Length 41.0
Proximal anteroposterior breadth 4.2
Tibiotarsus Breadth at midshaft 3.1
Distal anteroposterior breadth 4.2
Length 20.0
Proximal breadth 5.0
Tarsometatarsus Distal breadth 4.5
MtV length 8.0
Approximate length of phalanx 1 5.0
Pes digit I Ungual length 5.5
Phalanx 1 length 7.0
Pes digit II Phalanx 2 length 5.2
Ungual length 6.5
Phalanx 1 length 6.0
Phalanx 2 length 5.0
Pes digit III Phalanx 3 length 5.0
Ungual length 8.0
Phalanx 1 length 4.0
Phalanx 2 length 3.0
Pes digit IV Phalanx 3 length 3.0
Phalanx 4 length 4.5
Ungual length 7.0
Length 2.3
Ilium Breadth at midsection 5.2
Acetabulum anteroposterior breadth 4.0
Length 13.3
Proximal breadth 5.4
Ischium Distal breadth 5.0
Dorsal process height 3.0
Length 22.0
Pubis Proximal breadth 3.0
Distal breadth 2.3

individually clawed manus digits, unfused metacarpals, and hindlimb with unfused tarsals. The position of these two genera in the origin and evolution of birds is quite evident, and although both are biogeographically distinct and their remains have been subjected to different taphonomic processes, they share noticeable morphological characters that indicate that after birds diverged from their reptilian ancestors, their evolution occurred at a modest rate in the Late Jurassic. This is particularly noticeable in the large quantity of specimens excavated from the Early Cretaceous of China (Hou and Liu, 1984; Hou and Zhang, 1993; Hou, 1994; Hou, 1995; Hou et al., 1995). Although the evolutionary level of Archaeopteryx and Confuciusornis is extremely basal, they are distinct from each other in numerous characters as enumerated below:

1. Archaeopteryx bears numerous acute maxillary and dentary teeth while Confuciusornis is basically edentulous with the exception of an extremely small amount of premaxillary teeth.

2. The rostral region of Archaeopteryx possesses an abundant dentition and lacks grooving or striations that would indicate a keratinous beak. On Confuciusornis the premaxilla and anterior maxilla are grooved and striated, which distinctly indicate a keratinous beak. Supplemental dentition is absent. Thus, Confuciusornis represents the most archaic beaked bird.

3. The anterior frontal on Archaeopteryx is relatively broad, whereas on Confuciusornis it is extremely narrow and expands only posterodorsal to the orbit.

4. Controversy exists regarding the nature of the squamosal on Archaeopteryx but on Confuciusornis the squamosal is well developed with a distinct quadrate-squamosal contact.

5. Confuciusornis possesses a distinct prefrontal but the presence of this element on Archaeopteryx is controversial.

6. The posterior parietal crest of Archaeopteryx is not as high and broad as that of Confuciusornis.

7. On Confuciusornis, a long process extends posteriorly from the apex of the supraoccipital and is fused to the ventromedial crest of the parietal. This construction is distinct from Archaeopteryx.

8. Confuciusornis possesses a large exoccipital tympanic bulla which may represent one of the fully pneumaticized neurocranial chambers. This is absent on Archaeopteryx.

9. The lacrimal of Archaeopteryx has a well developed anterior branch but on Confuciusornis the lacrimal has a well developed posteriorly oblique branch.

10. The proximal humerus of Archaeopteryx is narrow and long but in Confuciusornis is broad and transversely expanded; its breadth is more than half the length of the shaft and a distinct pneumatic foramen is present.

11. On the Confuciusornis manus, the first digit’s ungual phalanx is extremely long and exceeds the length of the first phalanx, but on Archaeopteryx this same element does not exceed the length of the first phalanx.

12. The distal ischium on Archaeopteryx is bifurcated but in Confuciusornis it is not bifurcated and lies at an oblique angle.

13. The distal pubes of Archaeopteryx are broad, spacious, booted, and fused to each other but on Confuciusornis the distal pubes are not expanded, the termini are rounded, and not fused.

14. Archaeopteryx has a straight dorsal margin of the ilium but on Confuciusornis this margin is crescentic.

15. Archaeopteryx lacks a patella but Confuciusornis possess an elliptical patella.

16. Archaeopteryx retains three distinct tarsals but Confuciusornis retains only one which is fused neither to the tibiotarsus nor tarsometatarsus.

17. Although there are a number of Archaeopteryx specimens only flight feathers are represented and torso feathers are extremely rare. Confuciusornis not only preserves flight feathers, but also abundant torso feathers, including down feathers.

A complete list of discrepant characters is not provided above although the characters enumerated are sufficient to justify the extreme distinctions between the two genera. The derived characters of Confuciusornis include the presence of a keratinous beak, absence of dentition, and presence of a pneumatic foramen on the humerus. A beaked primitive bird in the Late Jurassic is extremely noteworthy, because it represents the most archaic beaked bird, it indicates that the presence of a keratinous beak is a significant character in the early evolution of birds, and that it is a necessary implement in avian feeding. The reduction or loss of dentition corresponds to the presence of a beak. Following the presence of the keratinous beak and loss of the dentition, the digestive system of birds probably began modification, including the specialization of the stomach and the ingestion of gastroliths for grinding foodstuffs. This would replace the function of the dentition.

The specialization of the humerus in Confuciusornis indicates a correspondingly adaptive pectoral girdle. The presence of the pneumatic foramen is related to the proximal transverse expansion and represents an attempt to decrease skeletal weight while increasing the ability for flight, thereupon increasing the strength of the humerus and modifying the pectoral girdle.

The presence of flight feathers and secondary torso feathers (including down) completes the distribution of feathers on the avian torso in the Late Jurassic. The absence of these feathers on Archaeopteryx may be related to the depositional environment. Comparison of these two genera leads to the following conclusions: There are noticeable distinctions between the two genera although fundamentally they are morphologically similar. The two genera are biogeographically distinct, but together with the documentation of an Archaeopteryx forelimb in Korea, indicate that there are at least three localities that produce Late Jurassic avian data. In the Late Jurassic the rate of avian evolution is quite distinct from that of the Early Cretaceous.

The current archaic avian data at hand is restricted, such that additional better preserved specimens are required to advance the comparative analysis of Jurassic and Cretaceous birds, confirm avian evolutionary patterns, allow further comprehension of biogeographical distribution in the Jurassic, assist in the comprehension of avian evolution and diversification, and provide better data for determining the origin of the class Aves. Current data implies that there was a tendency for avian lineages to radiate in the Jurassic.

New data on Confuciusornis In the fall of 1995, Yueming Hu and others traveled to the Jianshangou region of Beipiao, Liaoning, to initiate investigations into Mesozoic vertebrates. He obtained a fossil bird specimen from one of the local residents of the small village of Sihetun near Jianshangou, on the western Chaoyang-Beipiao thoroughfare, and thereupon immediately telephoned the author of this monograph. Hou immediately traveled to the village to find three households which had collected a total of four Confuciusornis specimens, in addition to an exceptionally valuable different and smaller specimen which was excavated from the same stratigraphic horizon as Confuciusornis. The locals then provided Hou with a tour of the fossil producing locality for stratigraphic clarification, where he was astonished to find a large contingent of local inhabitants in the process of excavating fossils along a single productive layer. As a scientist, Hou had no authority to interfere in the activities of the local populace and as such found it best to notify the appropriate office of the local authorities about the presence of a paleontological heritage site. The several specimens of Confuciusornis excavated in 1994 were incompletely preserved and were derived from different localities. The new specimens are relatively well preserved and consist of three specimens (including a single feather) produced from equivalent stratigraphic units which produced the skull and forelimb that represent the type. These sediments of mudstones and marls are consistent with those of the type locality.

After the initial publication that erected the genus Confuciusornis, there was extensive interest and strong reactions from paleontologists within and outside of China. These new specimens allow a more complete and supplemental description of the morphology of Confuciusornis, particularly in the postcrania and torso feathers, in addition to allowing a more complete comprehension and comparison to Archaeopteryx, and increasing the understanding of early avian radiation. These new specimens indicate that the genesis of the class Aves was neither in the Late Jurassic nor was this time period the first phase of avian evolution. The new specimens indicate a specialization of forelimb claws and the presence of a large sternum, and confirm the large proximal pneumatic foramen on the humerus, indicating that Confuciusornis was on the main path toward active flight. The presence of Confuciusornis supplants the traditionally recognized position of Archaeopteryx as the sole representative of the archaic condition. Its morphology forces workers to renew their opinions about primitive birds, avian origins, and developmental directions.

Confuciusornis sanctus Hou, Zhou, Gu, and Zhang, 1995 Specimens: A single individual lacking its right forelimb, hindlimb, and tail (V110304). Locality and stratigraphic position: Basal light gray mudstones of the Late Jurassic Yixian Fm., Sihetun Village, near the town of Shangyuanzhen, Beipiao (Beipiaoshih) Municipality, Liaoning Province.

Description

The skull has been subjected to dorsoventral and slight left lateral compressional distortion. The anterior left side is overlain by the right side and a large portion of the posterior frontal has been crushed such that suture lines are difficult to distinguish. However, clearly preserved on the right side are elements including the mandible, orbital elements, premaxilla, and nasals, which are all morphologically consistent with those of the type and clarify several details. The premaxilla is relatively large with conspicuous, regularly patterned, relatively deep striations associated with relatively thick grooves on its margin.

The anterior end of the premaxilla is not very acute and the nasal process of the premaxilla is not elongated, which are characters consistent with the type, and represent plesiomorphic characters for the genus. The premaxillary ramus is relatively long and high and its posterior margin contacts the maxilla with a long and oblique suture line. The small external nares are bounded anterolaterally by the premaxilla and posteriorly by the anterolateral portion of the nasals. The small size of these nares is in distinct contrast to the large and elongated external nares of Archaeopteryx. The anterior maxilla is high and its vertical posterior margin forms the anterior wall of the antorbital fenestra. The posteriorly extended ramus of the maxilla becomes narrow and forms the ventral wall of the antorbital fenestra. Striations on the maxilla are not as conspicuous as on the premaxilla and grooving is also light. The terminus of the maxillary ramus is in contact with the jugal.

Further observations are not possible due to compression from the mandible. Here, it is noteworthy to document a triangular element that forms the anterior wall of the antorbital fenestra between the posterolateral margin of the premaxillary nasal process and the posterior margin of the longitudinal branch of the maxilla. The presence of this element is confirmed on the type and hypodigm and it is determined to be a septomaxilla, an element that is located in various positions on the skulls of diapsids, including “thecodonts,” the Mesosauria, and therapsids. On Confuciusornis this element more closely resembles its homologs on the thecodonts. If this element is indeed a septomaxilla, its presence would be exceptionally significant in determining which clade of reptile gave rise to Aves and would lend credence to the theory that Aves originated from the thecodont reptiles. The nasals are completely preserved and are consistent with those of the type, being a pair of linear plates that are not transversely expanded. Surficially, they are smooth, glossy, and lack any conspicuous pitting ornamentation. Posterolaterally, the nasals contact two small triangular postorbitals and anteriorly they contact the premaxilla, where there are a few short striations.

The interorbital region, where the posterior nasals contact the anterior frontal, is extremely narrow. The nasals form the medial wall of the extremely large, elliptical antorbital fenestrae and occupy the medial surface of the palate, being isolated from the orbit only by the relatively large descending branch of the lacrimal. The lacrimal is an oblique, elongated, and triangular element. Further description of it is impossible due to being compressed against the sclerotic ring. The sclerotic ring is poorly preserved on the type but is completely preserved on the new specim en. It is surprising to note the thickness of the individual plates as opposed to the thin plates on extant taxa. Moreover, some of the plates are relatively large, providing extremely good protection for the eye. This specimen is comparable to the Eichstatt Archaeopteryx specimen, which also possesses a
relatively well preserved sclerotic ring.

The cranium posterior to the posterior frontal is completely crushed and suture lines are indistinct, although the frontal appears to be consistent with the type; its anterior interorbital portion is extremely narrow and expands abruptly at the posterior wall of the orbit. Its posterior contact with the parietal is indistinct because the parietal is also extremely damaged. Cranial contact with the atlas is tight.
In lateral perspective, the main body of the quadrate is extremely distinct and is in articulation with the mandible. Its posterior margin is medially depressed and its posterodorsal process projects to contact the squamosal, although it has become displaced. The squamosal extends posterolaterally with its medial side overlapping the quadrate process. As on the type specimen, the squamosal is relatively well developed.

Most recently, Elzanowski and Wellnhofer (1996) reevaluated the skull of Archaeopteryx, confirming that the squamosal is a thickened plate, which is quite distinct from Confuciusornis on which it is a rather expanded cranial element. The anterior quadrate has a folded surface. Its anterior process has been compressed upon and interlocks with the pterygoid. The contact of the anteroventral angle of the quadrate with the quadratojugal has also been displaced and the gracile quadratojugal is exposed ventral to the posterior end of the mandible. The right mandible on the new specimen is extremely well preserved. Fusion is incomplete at the symphysis as a suture line is still quite conspicuous. A well developed mandibular fenestra is present posteriorly. The retroarticular process does not extend posteriorly and instead is directly vertical, has a relatively flat ventral margin, and its anterodorsal margin ascends extensively to form a convex arc, resembling a design for intensive food aquisition. Anterolaterally, the mandible is striated but posteriorly it is relatively smooth and glossy. The dentary forms the lateral wall of the mandible, with its anterodorsal margin as a convex arc and the lingual surface being concave.

In medial perspective, the angular extends anteriorly to become tightly fused to the lateral side of the prearticular and on the medial side of the dentary extends posteriorly to the posterior mandibular fenestra to contact the medial margin of the articular. The surangular is robust, long, and represents the second largest element on the mandible next to the dentary. It forms the posteroventral margin of the mandible and extends directly anteriorly to the anterior portion of the dentary, while posteriorly it extends to the posterior end of the mandible to contact the dorsomedial margin of the articular.

The prearticular lies between the angular and surangular as an elongated plate, forms the ventral margin of the mandibular fenestra, and its posterior margin contacts the anterolateral side of the articular. The articular is relatively compressed, but tall and thick, and represents the most posterior element on the mandible. The glenoid is shallow with distinct lateral pitting, in distinct contrast to Archaeopteryx. A short and small vertical element, identified as the coronoid, is present at the anterior mandibular fenestra, although it does not project dorsally to surpass the surangular and form a process. At the medial midsection of the mandible there is a large splenial which lies medial to the surangular as an extremely thin plate that extends directly anteriorly. It is distinct from the splenial of Archaeopteryx which is triangular, whereas on this specimen it is an elongated crescent.

The vertebral column of the new specimen is incomplete although it preserves more elements than on the type specimen, including cervicals, a single thoracic, and caudal vertebrae. The cervicals are basically complete, the first (atlas) being the most distinctive of the vertebrae and covering the occipital condyle. Posterior to the second cervical, the neural arches have become separated from the centra to form a single groove running down probably eight to nine vertebrae. Neural spines are all relatively low, a ventral keel is present on the centra which are rather short and are not heterocoelous, the neural canal is relatively large, and diapophyses unite to form a haemal canal, although posterior cervicals lack this structure. Bicapitate and robust cervical ribs are rather conspicuous, initiate on the fourth cervical and elongate posteriorly along the column. There is a single anterior thoracic vertebra with relatively large diapophyses, a relatively long ventral keel, and its neural canal is also relatively large. A single caudal vertebra is associated with the distal end of the tarsometatarsus, with a circular centrum, extremely well developed diapophyses, and an extremely low neural spine.

Only the left scapula and coracoid are missing on the pectoral girdle; the remaining elements therefore provide reliable data toward a unilateral comparison to Archaeopteryx. The furcula is completely represented on this specimen and although it has been shifted in position, where it lies at the rostral region of the skull, it is completely undamaged. Its size and morphology are consistent with that of Archaeopteryx, with the exception of the clavicle branches which are distinctly short, thick and flatly broadened. At the proximal end of each branch there is an oblique and elongated depression and at the termini there are coarsened articular condyles. At the proximal articular condyles, the clavicle branch is reduced to become slender. The distance between the clavicle termini is 27 mm, which is 5 mm shorter than the London Archaeopteryx specimen. The right scapula and coracoid are preserved in natural configuration which is extremely benneficial toward the comprehension of the pectoral girdle.

The preservation on Archaeopteryx is similar in that on the vast majority of specimens the scapula and coracoid are also preserved in articulation. These elements on both genera are basically similar. The contact of the scapula and coracoid is medial, there is a large rounded glenoid fossa, and the scapula is long and plate-shaped with a reduced distal end. The proximal coracoid is large and expansive with a concave dorsal surface and lateral surface that is expanded to form a crescentic-shaped process that extends distally to become extremely thin. Its posterior side has a groove for contact with the sternum and its terminus is concave. The coracoid morphology of Confuciusornis is quite distinct from all those on the Early Cretaceous specimens and more closely approaches the morphology of its reptilian ancestors. On the type a sternum is absent, but on the new specimen, the sternum is completely preserved. Its morphology is unexpected, being particularly autapomorphic. It is much more well developed than that of Archaeopteryx and is also distinct from the Early Cretaceous specimens. Both a carina and posterior process are absent; its anterior end is rounded and descends at its sides while laterally becoming expansive and spacious.

On each side there extends a short and relatively expansive “anterior process.” The terminus of this process does not protrude to form a vertical notched line and at the posterior portion of the lateral process it is very slightly expanded. The midline of the sternum is slightly posteriorly projected, thereupon making the terminus a blunt triangle. Although a fully formed carina is absent, there is still a very slight projection upon the midline of the body and at each anterior and posterior end there is a circular depression which may facilitate an incipient increase of sternal musculature. The right side of this specimen is overlain by the right tibiotarsus prohibiting a complete description. The morphology of this primitive bird’s carina is unique. Ribs are relatively well preserved. Cervicals, thoracic, and gastric ribs are preserved randomly on the block. Cervical ribs gradually elongate posteriorly along the column. In general morphology the proximal ends are rather expanded and distinctly bicapitate. Thoracic ribs are distinctly elongated, gracile, and slightly curved with more narrow and constricted proximal ends. Gastric ribs are slender and rounded with slightly thickened proximal ends and extremely thin distal ends. They are systematically aligned and intersect each other to compose a latticed yoke configuration.

The forearm digits are incomplete on the type specimen, but on the new specimen the left forearm is well preserved with a large amount of feathers, allowing a reconstruction of the forearm and a comprehension of the flight feather configuration. On the right side only the proximal humerus is preserved. The left humerus is complete with a robust but relatively short shaft and an elliptical proximal pneumatic foramen which is surrounded by extremely thin bone which gradually thickens laterally as does the shaft. Thus the proximal end is expansive and extremely thickened, with a large and pronounced medial tuberosity, a feature that is obscured on the type. Its counterpart, the lateral tuberosity, is relatively small and positioned more distally, with a thickened posterior wall. There is a low trough at the lateral side of the pneumatic foramen such that the most depressed region of the pneumatic foramen is at its proximal end, clearly developing the structure of pneumatization. The proximomedial side of the pneumatization has a projected beam that extends directly to the medial tuberosity and then posteriorly fuses to the shaft. The shaft is extremely straight and lacks the curvature found in later avian taxa. Approaching the distal end, the shaft expands slightly. The medial condyle is large but its condylar process is small, whereas the lateral condylar process is large.

The radius and ulna are slightly shorter than the humerus. The ulna is consistent with the type, displaying a slight degree of curvature, its proximal end is not very large, and there is only a shallow olecranon fossa for articulation with the medial condyle of the humerus. Distally there is a shallow carpal articular facet, as in Archaeopteryx but not in later avian taxa, which have a large trochlea for the carpometacarpus. The radius is slightly thinner than the ulna and has a straight shaft and a humeral cotyle on its proximal end. Its distal end is expanded with a carpal articular facet.

On extant avian taxa, the articulation of the humerus with the radius-ulna is quite distinct from that of archaic taxa, for on the former the proximal ulna has a medial and lateral cotyle for articulation with the humeral condyles, and on the radius the proximal humeral cotyle combines with the lateral cotyle of the ulna to articulate with the lateral condyle of the humrus. In Confuciusornis and Archaeopteryx the proximal ulna lacks medial and lateral cotyles and as such articulation occurs solely with the medial condyle of the humerus. This feature is a plesiomorphic character reflecting an inability to fly, and although the ulna maintains secondary flight feathers upon it, the humerus still retains the articular relationship derived from reptilian ancestors which lacked adequate force for propulsion.

Carpal definitions are more distinct on the new specimen than on the type. The type maintains a large and irregularly shaped carpal on the ulnar side of the manus, while on the radial side there is a single columnar shaped carpal. One side is rounded and smooth while the other side is relatively compressed. A third small carpal lies between these two which, on the type, is a thickened plate but on the new specimen the third carpal more closely resembles an ellipse located at the proximolateral side of the second metacarpal. In summary, Confuciusornis and Archaeopteryx are consistent in the presence of three carpals. The metacarpals on the new specimen are morphologically consistent with the type. McI is not elongated but is particularly broadened with a thick lateral margin but approaching the margin of McII it becomes extremely thin.

A large concave surface lies on the midshaft, the morphology of the entire element is slightly irregular, a distal facet is present for articulation with digit I, and the lateral side retains a muscular-tendinal depression. Metacarpal II is the most robust and longest in the series, has an expanded proximal end with a convex articular surface, a broad and flattened shaft, and an extremely large articular surface for digit II. Metacarpal III is more slender than McII but of equivalent length, its proximal end is not expanded and instead is constricted. Although the first manus digit has only one phalanx aside from the ungual, it is relatively long, possesses lateral ligament fossae, and articulates with the large and scimitar-shaped ungual with a well developed broad articular facet.

Proximally, the articular facet for McI is expanded with a relatively large articular facet. The ungual claw of the first digit is the largest among the three. Its keratinous sheath is well preserved and 12 mm longer than the ungual. A bony cushion lies at the base of the ungual associated with a deep concave facet for articulation with the first phalanx, lateral grooves are present that extend nearly to the apex of the claw, and its curvature is nearly 90°.

The first phalanx on manus digit II is as extremely broadened as McII, both sides of the shaft are thick, and at its midsection there is a longitudinally elongated depression which matches the condition of McII. Both these characters are incipient adaptations for flight. On Archaeopteryx, the second metacarpal and digit II are the most robust elements on the manus but do not exhibit the same conspicuous modifications for flight as in Confuciusornis. The second phalanx of digit II in Confuciusornis is longer than the first phalanx, is not as broad, and gradually becomes slender distally. The ungual of digit II has shifted to a position approaching the ungual of digit I. It is short, small, robust, has only a slight amount of curvature, and is obviously evolving into a phase of gradual reduction, being directly related to strengthening the ability for flight. In Archaeopteryx, the ungual on digit II is extremely large and on some specimens is larger, stronger, and more curved than the ungual on digit I. Generally in Archaeopteryx, the unguals on digits I and II are equivalent in size and the ungual on digit III is the smallest. This is in distinct contrast to the condition in Confuciusornis and confirms that the second digit on Archaeopteryx, although being capable in assisting in flight, has not been modified distinctly from its morphology for scaling trees and has not become distinctly adapted for flight. Therefore, it is concluded that Archaeopteryx fundamentally did not possess the ability for flight, and although it possessed feathers, its behavior lay in the mode of scaling trees with the use of its claws and gliding between
the foliage. There is a large amount of data that confirms that Archaeopteryx was a terminal side branch of Aves.

Digit III of Confuciusornis has three phalanges, all of which are relatively short and slender. Its ungual phalanx is much larger than that on digit II and resembles the ungual on digit I, also has an approximate 90° curvature, and laterally there are even broader lateral ligament grooves. This indicates that when Confuciusornis was scaling a tree, it principally utilized digits I and III for a handhold as these claws were obviously strengthened, which alleviated digit II for facilitating the entire function of flight, and increased the ability for gliding between trees. On the new specimen the second phalanx of digit II has slight curvature. This is also confirmed by observations of several other specimens and is also recognized as a modification for the adaptability of flight. When the author and his colleagues were in the process of amassing the large collection of Jurassic primitive avian specimens, a large amount of feathers were recovered both attached to individual skeletons, as individual specimens, or in an aligned series. These specimens represent torso feathers, primary feathers, down feathers, and flight feathers. There is still no adequate microscopic analytical technique for the study of fossil feather impressions and thus detailed observations on the microstructure of various types of feathers cannot be conducted. This is extremely regrettable because the microstructure of the feather is directly related to the ability to fly.

If it were possible to study the feathers of Confuciusornis microscopically, it would then be possible to conduct comparisons with the feathers of both extant taxa in addition to Archaeopteryx, which would lead to the recognition of the evolutionary phases of feathers in addition to determining the genesis of genuine flight feathers. Currently, from the perspective of general morphology, it may only be confirmed that Archaeopteryx and Confuciusornis feathers have extremely short barbs, and thus the entire feather is distinctly narrow and long, unlike those on extant taxa which are broad and thick. It is possible that both Archaeopteryx and Confuciusornis lacked the complex interlocking construction present in extant taxa and this is the primary reason why they did not possess a true ability to fly.

Among the hindlimb elements is a pair of tibiotarsi lying posterior to the sternum. The tibiotarsus is thin and long, its shaft is straight, and the proximal articular surface is convex. The distal condyles are slightly anteriorly projected; the distal end resembles the proximal end in having a convex articular surface, and an intercondylar groove is extremely narrow. At the distal end of the left tibiotarsus there is an incomplete femur overlying the proximal end of the tarsometatarsus. The femur is more robust than the tarsometatarsus, its proximal end is greatly inflated, but its head is missing as are the distal condyles. A right fibula is preserved which is extremely long, has a well developed proximal end, and a shaft which gradually attenuates to become extremely slender at its distal end. Only the proximal end of the tarsometatarsus is preserved overlain by the femur, which prohibits the description of its morphology, but in cross-sectional view it appears to have been composed of three unfused tarsals.

Skeletal measurements of the new specimen of Confuciusornis (mm).
Cranial length 73.0
Greatest posterior breadth of cranium 40.0
Antorbital fenestra length 6.5
Antorbital fenestra breadth 11.0
Mandible length 62.0
Preserved cervicals length 60.0
Scapula length 52.0
Coracoid length 28.0
Humerus length 61.0
Proximal humerus breadth 31.0
Pneumatic foramen length 11.0
Distal humerus breadth 15.5
Ulna length 55.0
Radius length 52.0
Metacarpal I length 11.0
Metacarpal I breadth 6.5
Digit I length 19.0
Digit I ungual length 16.5
Metacarpal II length 32.5
Metacarpal II breadth 6.0
Digit II length 46.0
Digit II ungual length 8.0
Metacarpal III length 31.0
Digit III length 33.0
Digit III ungual length 21.0
Furcula length 21.0
Furcula shaft breadth 6.0
Furcula breadth at suture 5.5
Sternum length 45.0
Sternum anterior process breadth 41.0
Sternum distal breadth 42.0
Tibiotarsus length 66.0
Tibiotarsus proximal breadth 10.0
Tibiotarsus distal breadth 9.0
Fibula length 35.0
Fibula proximal breadth 6.0
Femur preserved length 43.0

Discussion

The new specimen not only supplements significant morphological characters of Confuciusornis, but also allows a more comprehensive understanding of the nature of the genus, and provides more substantial characters to compare with Archaeopteryx. Its scapula is extremely similar to the latter, its furcula appears to be a little more primitive, and its clavicle branches are extremely short, all of which restrict Confuciusornis to a streamlined, rather short torso. The scapula-coracoid morphology is extremely close to that of Archaeopteryx; both elements are tightly articulated, and resemble the morphology associated with the Reptilia. Although the coracoid morphology of Confuciusornis is distinct from that of reptiles, it is always associated with the scapula, which is a plesiomorphic condition derived from the most recent ancestor of Confuciusornis and Archaeopteryx. The coracoid morphology of both archaic genera is vastly distinct from those of modern avian taxa. The new specimen displays a complete forearm associated with primary feather impressions. Combined with the particular modifications of the second metacarpal, digit II, and ungual of digit II, it provides comprehension of specific modifications for the origin of flight, as prior to the discovery of this specimen only vague hypotheses were formulated regarding feather development and forelimb modifica