On the brain of a primitive bird from the Upper Cretaceous of European Russia

Abstract

Cerebavis cenomanica gen. et sp. nov. from the Middle Cenomanian of the Volgograd Region (Russia) is described based on a brain mold. The brain
of Cerebavis is characterized by a mosaic combination of primitive and advanced features. The brain weight is estimated as approximately
1 g. The cerebrum is relatively very large, but lacks sulci. The brain mold has long olfactory lobes with large olfactory
bulbs, a well-developed epiphysis, and a parietal organ. The auditory tubercles on the dorsal surface of the midbrain are
well developed. The optical lobes are located under the auditory lobes, caudoventral to the cerebral hemispheres. The cerebellum
is not preserved, but its imprints just behind the midbrain suggest that it was probably relatively small and extended dorsoventrally.
The brain of Cerebavis is similar in some features to that of Archaeopteryx, but is substantially more advanced and more specialized. Cerebavis is similar to living ornithurine birds in the large cerebral hemispheres, but differs in the absence of a well-developed
neostriatum, the presence of excessively developed olfactory lobes, and in the pattern of the midbrain. Thus, senses of smell,
eyesight, and hearing were well developed in Cerebavis. It could have been equally active in the afternoon and at night. The unique brain design demonstrated by Cerebavis has not been repeated in subsequent evolution. It provides evidence for a wide diversity of feathered creatures in the past.
Cerebavis probably belongs to the Enantiornithes.

Full text

655
ISSN 0031-0301, Paleontological Journal, 2006, Vol. 40, No. 6, pp. 655–667. © Pleiades Publishing, Inc., 2006.
Original Russian Text © E.N. Kurochkin, S.V. Saveliev, A.A. Postnov, E.M. Pervushov, E.V. Popov, 2006, published in Paleontologicheskii Zhurnal, 2006, No. 6, pp. 69–80.
INTRODUCTION
The nervous system of birds and dinosaurs is of spe-
cial evolutionary and zoopsychological interest, since
the brain of extant ornithurine birds is extremely spe-
cialized, while the data on the diversity of feathered
dinosaurs and extinct birds has increased considerably
during the last years (Chiappe and Witmer, 2002). Even
an isolated endocranial cast provides important evi-
dence for the reconstruction of certain stages of the
complex problem of the origin of birds (Whetstone,
1983; Kurochkin, 2001, 2004; Alonso et al., 2004). For
this reason, the unique endocranial cast of the London
specimen of
Archaeopteryx
(BMNH, no. 37001),
housed in the Natural History Museum (London), was
repeatedly examined over the last 80 years (Edinger,
1926; de Beer, 1954; Jerison, 1968, 1973; Dechaseaux,
1968; Whetstone, 1983; Nieuwenhuys, 1998; Alonso
et al., 2004).
The present study describes a natural mold of bird’s
brain of Cenomanian Age from the Melovatka-3 local-
ity in the Zhirnovskii District of the Volgograd Region
of Russia.
LOCALITY MELOVATKA-3
The Melovatka-3 locality is situated on the right
slope of ancient erosion terrace of the Medveditsa
River, one of the largest tributaries of the Don River,
50°52
′
50.7
″
N,
44°44
′
10.9
″
E (WGS 84) (Fig. 1). It was
repeatedly visited by the authors of this study from
Saratov and by E.N. Kurochkin. The right (western)
slope of the Medveditsa River valley is a high erosion
terrace cut by short gullies varying in depth and extend-
ing towards the floodplain of the river. A section of
marine Turonian and Cenomanian deposits more than
70 m thick outcrops in the gullies of the terrace. The
Turonian beds are white and grayish yellow marls (Per-
vushov et al., 1999b); the Cenomanian beds are mostly
white and yellowish quartz-glauconitic and quartz
sands, supplemented in the lower layers by aleurites
(Pervushov et al., 1999a). The Middle and Upper Cen-
omanian deposits of this region are represented by the
Melovatskaya Formation spreading on the right bank of
the Volga Region near Penza, Saratov, and Volgograd
(Olfer’ev and Alekseev, 2005). The upper part of the
Cenomanian section in the Medveditsa River valley
On the Brain of a Primitive Bird from the Upper Cretaceous
of European Russia
E. N. Kurochkin
a
, S. V. Saveliev
b
, A. A. Postnov
c
, E. M. Pervushov
d
, and E. V. Popov
d
a
Paleontological Institute, Russian Academy of Sciences, Profsoyuznaya ul. 123, Moscow, 117997 Russia
e-mail: enkur@paleo.ru
b
Scientific Research Institute of Human Morphology, Russian Academy of Medical Sciences,
ul. Tsyurupy 3, Moscow, 117418 Russia
c
Lebedev Physical Institute, Russian Academy of Sciences, Leninskii pr. 53, Moscow, 117924 Russia
d
Saratov State University, Astrakhanskaya ul. 83, Saratov, 410012 Russia
Received December 27, 2005
Abstract
—
Cerebavis cenomanica
gen. et sp. nov. from the Middle Cenomanian of the Volgograd Region (Rus-
sia) is described based on a brain mold. The brain of
Cerebavis
is characterized by a mosaic combination of
primitive and advanced features. The brain weight is estimated as approximately 1 g. The cerebrum is relatively
very large, but lacks sulci. The brain mold has long olfactory lobes with large olfactory bulbs, a well-developed
epiphysis, and a parietal organ. The auditory tubercles on the dorsal surface of the midbrain are well developed.
The optical lobes are located under the auditory lobes, caudoventral to the cerebral hemispheres. The cerebel-
lum is not preserved, but its imprints just behind the midbrain suggest that it was probably relatively small and
extended dorsoventrally. The brain of
Cerebavis
is similar in some features to that of
Archaeopteryx
, but is sub-
stantially more advanced and more specialized.
Cerebavis
is similar to living ornithurine birds in the large cere-
bral hemispheres, but differs in the absence of a well-developed neostriatum, the presence of excessively devel-
oped olfactory lobes, and in the pattern of the midbrain. Thus, senses of smell, eyesight, and hearing were well
developed in
Cerebavis
. It could have been equally active in the afternoon and at night. The unique brain design
demonstrated by
Cerebavis
has not been repeated in subsequent evolution. It provides evidence for a wide
diversity of feathered creatures in the past.
Cerebavis
probably belongs to the Enantiornithes.
DOI:
10.1134/S0031030106060086
Key words
: Cretaceous bird, Enantiornithes, fossil brain, avian brain mold, Late Cretaceous, Cenomanian, Vol-
gograd Region, Russia.

656
PALEONTOLOGICAL JOURNAL
Vol. 40
No. 6
2006
KUROCHKIN et al.
was dated Middle Cenomanian based mostly on the
mollusk and elasmobranch faunas (Pervushov et al.,
1999a). This section contains several levels of phos-
phate concretions of different generations and various
degrees of concentration. In the middle part of this
interval in the Melovatka-3 section, there are one or two
phosphorite horizons 0.3–0.6 m thick, which are rich in
phosphate concretions and phosphatized fossils of
marine invertebrates. The two phosphorite horizons are
separated in places by a yellowish sand interbed up to
1 m thick; however, they sometimes merge into a single
horizon. The lower interbed or part of the integrated
phosphorite horizon shows in places straight diagonal
bedding, which is manifested in the orientation of
almost flat pellets, pebbles, and pseudomorphs on
bivalve shells.
The phosphate concretions are irregular aggregates
and rounded pebbles, ranging in size from several mil-
limeters to 5–10 cm. They are dark brown or nearly
black, as the molds of numerous bivalves, gastropods, and,
more rarely, cephalopods (possibly
Schloenbachia
sp.).
Most of them are only represented by phosphate–cal-
cium–silica inner molds. The scarce brachiopods,
including the hinged
?Rhynchonella
sp. and hingeless
Lingula
sp., initially have a phosphate shell distin-
guished by a dark purplish color and retaining its struc-
ture in detail. Skeletal remains of siliceous sponges, the
Demospongiae (of the genera
Jerea, Siphonia, Selisco-
thon
, and
Phyllodermia
), are not numerous and usually
show poor preservation and varying degrees of phos-
phatization; this is attributable to redeposition. The
same is true of hingeless brachiopods and some mol-
lusks from the underlying deposits and sponges sub-
stantially damaged by bioerosion, which occurred
mostly before the main stage of fossilization. Bivalve
molds dominate in both number and diversity. They are
represented by species typical of Cenomanian seas of
the southern and southeastern Russian Plate:
Amphi-
donte conicum
(Sow.),
A. subconicum
(Glas.),
Crano-
cardium ventricosum
(Orb.),
C. constantii
(Orb.),
Pro-
tocardia hillama
(Sow.),
Hercodon aequilateralis
Sobetski,
Corbulamella elegans
(Sow.),
Panopea man-
dibula
(Sow.),
P. acutisulcata
(Desh.),
Neithea quin-
quecostata
(Sow.),
Chlamys hispida
(Goldf.),
Ento-
nium orbicularis
(Sow.),
Flaventia plana
(Sow.) (Per-
vushov et al., 1999a). Carbonate shells of mollusks are
almost completely dissolved and only small areas near
the hinge are occasionally preserved.
Vertebrate remains also occur in these phosphorite
horizons and are dominated by elasmobranch teeth.
100 km 012
km
(a) (b)
Medveditskii
Nizhnyaya
Dobrinka
Linevo
Medveditsa R.
12
3
N
Zhirnovsk
Melovatka
Volga R.
Saratov
Moscow
Russia
Kazakhstan
Ukraine
Belarus
Finland
Ural Mountains
Fig. 1.
Geographical position of the Melovatka-3 locality: (a) large-scale and (b) small-scale maps. Asterisks designate the Melo-
vatka-1, 2, and 3 localities.

PALEONTOLOGICAL JOURNAL
Vol. 40
No. 6
2006
ON THE BRAIN OF A PRIMITIVE BIRD FROM THE UPPER CRETACEOUS 657
The following elasmobranchs have been determined:
Eostriatolamia subulata
(Agassiz),
Cretolamna appen-
diculata
(Agassiz),
Odontaspis
sp.,
Megarhizodon
cf.
macrorhiza
(Cope, 1875),
Pseudoisurus denticulatus
(Glückman),
Palaeoanacorax volgensis
Glückman,
Paraorthacodus recurvus
(Trautschold),
Synechodus
dispar
(Reuss),
Squatina
sp., and
Polyacrodus
sp. In
addition to shark teeth, rare dental plates of chimaeras
were also found, including the edaphodontids
Edaph-
odon sedgwicki
Agassiz,
Ischyodus latus
Newton,
Elas-
modus sinzovi
Averianov, and
Elasmodectes kiprijan-
offi
(Nessov) and the rhinochimaerid
?Belgorodon
sp.
The locality has also yielded vertebrae and teeth of
bony fishes (including Enchodontidae), plesiosaurs
(Elasmosauridae indet.), and ichthyosaurs (
?Platyp-
terygius
sp.); large spiral coprolites of cartilaginous
fishes and small elongated coprolites of bony fishes are
common. Pieces of fossil wood are scarce, while ptero-
saurian and avian bone fragments are even less frequent
(Pervushov et al., 1999a; Averianov et al., 2005; sharks
were identified by V.K. Golubev and E.K. Sytchev-
skaya, PIN). The brain mold described below was also
found in these phosphorite horizons.
FOSSIL BRAIN
The specimen under study was found in 1993 by an
expedition of the Department of Historical Geology
and Paleontology of Saratov State University and trans-
ferred to the Paleontological Institute of the Russian
Academy of Sciences, Moscow (PIN). The specimen
was covered in places with spots and incrustations of
extremely durable phosphatized silica, which were very
difficult to remove by hand and hence complicated a
thorough examination. Therefore, it was initially mis-
taken for an avian skull. As the enclosing matrix was
removed with a pneumatic needle and the specimen
was prepared in the Museum of Natural History of the
University of Kansas (Lawrence, United States), it
became evident that we dealt with an almost complete
fossil brain. However, it remained unclear as to which
elements were fossilized bone structures of the skull
and which were fossilized brain. These questions were
resolved in the laboratory of the University of Antwerp
(Belgium), using a
Skyscan
1072 scanning x-ray micro-
tomograph, with the resolution 6
µ
m, 0.3–0.4 Å, 40–
80 kV. A series of 1022 x-ray sections of the specimen
was produced and recorded in the jpeg format. These
sections were used in the study of neurocranial relation-
ships of various regions of the fossil brain and bone
structures of the skull.
This raises the question as to how the soft cerebral sub-
stance could be preserved in fossilized condition. It was
shown that, in experiments in warm water saturated with
biogenes and minerals in alkaline medium of pH 8–9, bac-
terial fossilization of soft tissues of animals and plants
developed very quickly, in few hours or even minutes
(Gerasimenko et al., 1994; Abyzov et al., 2002). Simi-
lar conditions probably existed in the nearshore shallow
water near the islands of a Cenomanian sea that occu-
pied the southern area of the modern Russian Plain.
They provided a rapid postmortem siliceous–phosphate
fossilization of the brain of this bird, as was in the case
with demosponges and molluskan bodies, which pre-
served in detail the structure of soft tissues. Subse-
quently, the bony phosphate–calcium elements of its
skull were disrupted, while the siliceous–phosphate
brain mold was preserved. Chemical analysis with the
aid of a SEM in the Laboratory of Electron Microscopy
of PIN has shown that the specimen consists mostly of
calcium (40–50%), silicon (20–35%), and phosphorus
(15–20%), which are probably included in two main
minerals, i.e., calcium phosphate, Ca
3
(
PO
4
)
3
, and silica,
SiO
2
(
n
H
2
O
)
, which compose the nucleus of the brain.
SYSTEMATIC PALEONTOLOGY
SUBCLASS SAURIURAE HAECKEL, 1879
?INFRACLASS ENANTIORNITHES WALKER, 1981
Genus
Cerebavis
Kurochkin et Saveliev, gen. nov.
Etymology. From the Latin
cerebrum
(brain)
and
avis
(bird, feminine gender).
Type species.
Cerebavis cenomanica
sp. nov.
Diagnosis. Cerebral hemispheres rounded oval.
Olfactory tracts thick, with large olfactory bulbs. Inter-
hemispheric fissure shallow, parietal organ well pro-
nounced, located in pineal recess on caudal slope of
interhemispheric fissure; roof of midbrain with large
auditory tubercles, well-developed epiphysis (glandula
pinealis) located between auditory tubercles, optic
tubercles (lobi optici) located caudoventral to cerebral
hemispheres, not projecting laterally beyond them.
Middle part of parasphenoid rostrum swollen.
Species composition. Type species.
Comparison. It is possible and expedient to
compare the brain of
Cerebavis
with that of the London
specimen of
Archaeopteryx lithographica
von Meyer,
1861, which was repeatedly reconstructed and investi-
gated (Edinger, 1926; de Beer, 1954; Jerison, 1968,
1973; Alonso et al., 2004). The brain structure of enan-
tiornithines is not known.
Archaeopteryx
also belongs
to Sauriurae (Martin, 1983; Kurochkin, 1995, 1996,
2001). In general, the brain of
Cerebavis
is shorter,
higher, and wider mostly due to its large cerebral hemi-
spheres; the olfactory tract is longer, but equally thick-
ened (Pl. 6, figs. 1–6) compared to those of the endocast
of the London specimen of
Archaeopteryx
, the brain of
which is generally elongated (Figs. 2, 3). The hemi-
spheres are separated by a shallow interhemispheric fis-
sure, which is developed to approximately the same
extent as in
Archaeopteryx
. The caudal slope of the
hemispheres is substantially more abrupt than in
Archaeopteryx
. The parietal (parapineal) organ opens
in the interhemispheric fissure on the caudal slope in a
well-pronounced recess; it has also been recorded in
Archaeopteryx
, although a depression in this part of its

658
PALEONTOLOGICAL JOURNAL
Vol. 40
No. 6
2006
KUROCHKIN et al.
interhemispheric fissure is virtually absent. The caudo-
dorsolateral surface of the hemispheres has a character-
istic transverse fold (telencephalon recess), which is
also present in
Archaeopteryx
, although it is shallower
(Fig. 2; Pl. 6, figs. 3–5). The quadrigeminal plate (lam-
ina quadrigemina tecti) formed by two pairs of widely
spaced tubercles is clearly seen on the roof of the mid-
brain; the auditory (posterior) tubercles are on the dor-
sal surface of the midbrain, while the optic (anterior)
tubercles are on the lateral sides. A well-developed epi-
physis (pineal gland) is seen between the auditory
tubercles (Pl. 6, figs. 2, 4, 5), while it has not been
recorded on the dorsal side of the midbrain of
Archae-
opteryx
; the latter also lacks auditory tubercles in this
region (Fig. 2). The optic tubercles (optic lobes) are
smaller in both relative and absolute dimensions than in
Archaeopteryx
, terminate short of reaching the level of
the lateral surfaces of the cerebral hemispheres, and are
located more caudoventrally under these hemispheres.
In
Archaeopteryx
, the optic tubercles project out of the
lateral sides of the hemispheres, which is connected
with the smaller volume of the hemispheres, and lie
under the midbrain hemispheres, more caudally with
reference to the cerebral hemispheres.
Some cranial elements of
Cerebavis
are filled with
phosphate–silica matrix in the shape of voluminous for-
mations, which substantially differ from those of
Archaeopteryx
. The interorbital septum was high,
extending from the ventral base of the skull to the fron-
tals, with a small foramen in the center. The eminences
in the auditory region, following the outline of the squa-
mosal, prootic, and exoccipital are much larger in vol-
ume (Pl. 6, figs. 1, 4) than the small, narrow bones of
Archaeopteryx
(Fig. 2b). The formation following the
outline of the basiparasphenoid is a broad trapezoid
area (lamina parasphenoidalis), from under which an
expanded base of the parasphenoid deviates, passing
into the parasphenoid rostrum, which is swollen at the
base. Small eminences, which most likely correspond
to the bases of the basipterygoid processes, project ros-
trolaterally from the base of the parasphenoid.
R e m a r k s. Fossilized brain molds are extremely
scarce, particularly those of birds or birdlike creatures;
specimens of this kind have never been recorded in the
Mesozoic. Several phosphatized cerebral molds of true
Neornithes come from the Paleogene and Neogene
beds of Europe, they are rather similar to the brain of
extant birds (Dechaseaux, 1970; Mlíkovsk
, 1980).
The brain of Archaeopteryx was examined tomograph-
ically and based on endocranial casts. The brains of
Late Cretaceous Hesperornis regalis Marsh, 1872 and
Ichthyornis dispar Marsh, 1872 were reconstructed
based on the skull. The brain of Enaliornis barretti See-
ley, 1876 from the Albian of England was reconstructed
based on a medullary skull cavity.
Most of the fossil brain of Cerebavis is formed of
cerebral hemispheres (Pl. 6, figs. 2, 5; Pl. 7, figs. 11, 12).
The width of the hemispheres is somewhat greater than
the length, and they are separated from one another by
a distinct, but shallow interhemispheric fissure (Pl. 6,
fig. 4; Pl. 7, figs. 12, 13, 16, 17). Due to these broad lis-
sencephalic hemispheres and the extended trace of the
cerebellum, the general structural pattern of the brain of
Cerebavis on the dorsal surface is most similar to that
of extant Neornithes. The optic lobes of Cerebavis are
immersed under the base of the brain and are spaced
laterally almost as widely as in Neornithes. However,
certain essential differences are clearly pronounced.
The brain outline of Cerebavis is pear-shaped,
extended longitudinally to a somewhat greater extent
than in Neornithes, the brain of which is almost spher-
ical, short and high (Figs. 3b, 3c). A significant distinc-
tion of Cerebavis from Neornithes is the position of its
y´
po
rt
bo(o)
to(ot)
t(c)
rp(pg)
tm(ol)
m(cb)
vf(fl)
c(pc)
(a)
(b)
rp(pg) m(cb)
tm(ol)
vf(fl)
fm(fm)
c(pc)
rt
t(c) po
Fig. 2. Tomographic reconstruction of the brain of Archae-
opteryx lithographica von Meyer, 1861 (on Alonso et al.,
2004, modified; resolution no. 15055758 from Copyright
Clearance Center, Inc.): (a) dorsal and (b) caudal views.
Designations (in parentheses, the treatment of structures
after Alonso et al., 2004): bo(o)—bulbus olfactorius (olfac-
tory lobe), c(pc)—cerebellum (cerebellar prominence),
fm(fm)—foramen magnum, m(cb)—midbrain (cerebellum),
po—parietal organ, rp(pg)—pineal recess [pineal gland
(epiphysis)], rt—telencephalon recess, t(c)—telencephalon
[cerebrum (telencephalon)], tm(ol)—midbrain tectum
[optic lobes (metencephalic tectum)], to(ot)—olfactory
tract, and vf(fl)—vascular or lymphatic fascicle (floccular
lobe of the cerebellum). Scale bar, 10 mm.

PALEONTOLOGICAL JOURNAL Vol. 40 No. 6 2006
ON THE BRAIN OF A PRIMITIVE BIRD FROM THE UPPER CRETACEOUS 659
midbrain, which is located between the cerebral hemi-
spheres and the trace of the cerebellum, with the roof open
from the dorsal side (Pl. 6, figs. 4, 5; Pl. 7, figs. 16, 17).
The midbrain is described in detail below. Another dis-
tinctive feature of Cerebavis is its long and thick olfac-
tory tracts and large olfactory bulbs, which are only
represented by the caudal bases; however, they clearly
show large size of the entire bulbs. They are much larger
in relative size of the primary olfactory centers than those
of living Neornithes. This is evidence of the well-devel-
oped olfactory system of this extinct bird and its similar-
ity to extant archosaurs (crocodiles) (Figs. 3a, 3b).
As compared to the thoroughly examined brain cast
of Archaeopteryx housed in London (Edinger, 1926; de
Beer, 1954; Dechaseaux, 1968; Jerison, 1968; Whet-
stone, 1983; Nieuwenhuys, 1998; Alonso et al., 2004),
the brain of Cerebavis has shorter and more volumi-
nous cerebral hemispheres and longer olfactory tracts.
It is similar to the brain of Archaeopteryx in the pres-
ence of the telencephalon recess on the dorsal surface
in the caudal part of the hemispheres (Fig. 2; Pl. 6,
figs. 3, 5). This fold has been recorded in a number of
studies devoted to the brain of Archaeopteryx (but was
not mentioned by Alonso et al., 2004), although its
function remains uncertain (Dechaseaux, 1968; Nieu-
wenhuys, 1998). Apparently, this fold could have
developed in Archaeopteryx and Cerebavis as a result
of outstripping development of the neostriatum (Nieu-
wenhuys, 1998; Saveliev, 2001, 2005). The caudal
topological boundary of the hemispheres of Cerebavis
and Archaeopteryx is in line with the telencephalon
recess, as in the majority of living birds. In living birds,
III
III
C
TS
EP
MS
MT
TM
EP
MT
TS
EP
C
C
MS III
CEP
RT PO
T
T
RT
TM
BO
BO TO
HTM MT MT
TS
TM
H
TO
BO
T
MS MT
C
TS
TM
EP
MT
C
TS
TM
(a)
(b)
(c)
Fig. 3. Brain structure, laterally (on the left), with sagittal sections (on the right) in line with the midbrain through ventricle III:
(a) crocodile, (b) Cerebavis cenomanica sp. nov., and (c) pigeon. Designations: (III) third ventricle, (BO) olfactory lobe, (C) cere-
bellum, (EP) epiphysis, (H) hypophysis, (MS) midbrain, (MT) afterbrain (metencephalon), (PO) parietal organ, (RT) telencephalon
recess, (T) cerebrum (telencephalon), (TM) midbrain tectum, (TO) olfactory tract, and (TS) torus semicircularis (auditory tubercles).
Dash lines designate reconstructed regions of the brain of C. cenomanica sp. nov.

660
PALEONTOLOGICAL JOURNAL Vol. 40 No. 6 2006
KUROCHKIN et al.
the cranial surface of the cerebrum (telencephalon)
always has greater or lesser developed sulci of the cere-
brum (valleculae telencephali), while Cerebavis and,
more so, Archaeopteryx lack a trace of sulci. In addi-
tion, the interhemispheric fissure of Cerebavis is much
shallower than in Neornithes. Consequently, the cere-
brum of Cerebavis shows primitive structural charac-
ters; however, it is more similar to that of Neornithes
than Archaeopteryx.
The design of the midbrain of Cerebavis is espe-
cially interesting. It essentially differs from the brain of
both extant archosaurs (crocodiles) and Neornithes
(Fig. 3). The midbrain roof of Cerebavis is formed of
two pairs of widely spaced tubercles of the quadrigem-
inal plate. The auditory and optic tubercles are clearly
visible on the surface of the neurocranial mold (Pl. 6,
figs. 3–5) and in the x-ray sections through the caudal
region of the specimen (Pl. 7, figs. 13, 15). The two
pairs of tubercles on the midbrain roof are known in the
brain of mammals (Saveliev, 2005). However, in Cere-
bavis, they occupy positions that have not been
recorded in living birds or mammals. The auditory
tubercles of the midbrain roof are located atypically of
reptiles or mammals. They come onto the dorsal sur-
face of the midbrain between the cerebral hemispheres
and cerebellum (namely, a trace of the cerebellum in
the shape of a cavity in the specimen). The auditory
tubercles form two well-pronounced hemispheres, with
the epiphysis between them (Pl. 6, figs. 2, 5). The mid-
brain of Archaeopteryx occupies a similar position, if it
is thought to be represented by the optic tubercles and
two slightly convex, bilaterally segmented hemispheres
of the auditory tubercles located just behind the cere-
bral hemispheres, which were probably mistaken by
Alonso et al. (2004) for the cerebellum (Fig. 2). The
bilaterally segmented cerebellum, at least in extant rep-
tiles and birds, has not been recorded. It is also impos-
sible to concur with the treatment proposed by Alonso
et al. (2004) for two formations that substantially
project laterally and are located under the caudal bases
of the optic tubercles (Fig. 2). These researchers identi-
fied them as “floccular lobes of the cerebellum.” In this
case, the tomographic reconstruction probably dis-
played blood or lymphatic vascular fascicles. In addi-
tion, it should be noted that, in the paper cited, the optic
tubercles are incorrectly assigned to the afterbrain
(“metencephalic tectum”), whereas, in actual fact, they
always belong to the midbrain.
The cerebellum of Archaeopteryx, like that of extant
reptiles, was probably represented by a very small for-
mation that was designated by Alonso et al. (2004) as
“cerebellar prominence” located between the caudal
ends of the midbrain hemispheres. If our interpretation
is correct, it supports the idea of some researchers that
Archaeopteryx was only adapted for linear gliding from
one landing point to another rather than for a maneuver-
able and active flapping flight.
The tomographic reconstruction of the brain of
Archaeopteryx proposed by Alonso et al. (2004) lacks
auditory tubercles and epiphysis. We believe this fea-
tures are reconstructed correctly. The auditory tubercles
of Archaeopteryx were most likely immersed inside the
third ventricle, under the midbrain roof, as in extant
reptiles (Fig. 3a). However, Cerebavis and Archaeop-
teryx are similar in the dorsally open midbrain.
Cerebavis has a well-pronounced parietal organ
located rostral to the epiphysis, in the interhemispheric fis-
sure between the cerebral hemispheres (Pl. 6, figs. 2, 4). In
the reconstruction of the brain of Archaeopteryx
(Alonso et al., 2004), it is also distinct (Fig. 2), although
it was mistaken for the epiphysis [Alonso et al., 2004,
p. 667, “pineal gland (epiphysis)”].
The optic tubercles of the quadrigeminal plate of
Cerebavis are covered dorsally by the lateral margins of
the cerebral hemispheres; however, the position of
these tubercles is the same as in the majority of extant
birds, i.e., they are located caudoventral to the hemi-
spheres (Pl. 6, fig. 3; Pl. 7, fig. 13). They are identified
with certainty based on the position, while their relative
size is similar to that of the anterior tubercles of living
passerines. The optic tubercles of Archaeopteryx are
relatively and absolutely larger than those of Cerebavis
and are shifted substantially more caudally, so that their
rostral portions are only slightly covered dorsally by
the cerebral hemispheres, and the lateral surfaces
project somewhat out of the lateral margins of the
hemispheres (Fig. 2), as was shown by Alonso et al.
(2004). However, this is basically attributable to the rel-
atively smaller cerebral hemispheres.
The position of the auditory tubercles and their size
suggest that the brain of Cerebavis followed an unusual
path of evolution, which differed from that of ornithu-
rine birds and reptiles. In living reptiles, the midbrain
roof does not form a quadrigeminal plate, while the pri-
mary bulge of the auditory tubercles is enclosed in the
fold between the midbrain and cerebellum (Fig. 3a).
The midbrain roof of reptiles is almost entirely occu-
pied by the mesencephalic tectum, while the auditory
tubercles are not seen. Only in rare cases, for example,
in monitor lizards and crocodiles, the auditory tubercles
are seen as small tori. This region of the brain of
Archaeopteryx probably showed the same structural pat-
tern as in reptiles, judging from the previously published
brain reconstructions (de Beer, 1954; Dechaseaux, 1968;
Nieuwenhuys, 1998; Alonso et al., 2004). On the other
hand, some researchers believe that the cerebellum of
Archaeopteryx adjoins the caudal edge of the cerebral
hemispheres in the area where the epiphysis opens
(Alonso et al., 2004); as indicated above, this point of
view is in error. The possibility of such a high special-
ization of the brain of Archaeopteryx (Dechaseaux,
1968; Martin, 1995; Elzanowski and Wellnhofer, 1996;
Alonso et al., 2004) or birdlike dinosaurs, such as Inge-
nia yanshini Barsbold, 1981 (Osmólska, 2004), is
hardly probable. Apparently, this incorrect understand-

PALEONTOLOGICAL JOURNAL Vol. 40 No. 6 2006
ON THE BRAIN OF A PRIMITIVE BIRD FROM THE UPPER CRETACEOUS 661
ing of the brain of Archaeopteryx (Alonso et al., 2004)
is inherited from Jerison (1968, 1973), who mistook the
lobes adjoining posteriorly the cerebral hemispheres
for the cerebellum. Edinger (1926) and de Beer (1954)
believed that, in the London specimen, the cerebral
hemispheres are followed caudally by the midbrain,
without pronounced auditory tubercles; therefore, they
correctly regarded the brain of Archaeopteryx to be
closer to the reptilian than the avian brain. The cerebel-
lum of reptiles usually lacks transverse sulci and does
not approach the caudal edge of the cerebral hemi-
spheres.
In extant ornithurine birds, the auditory tubercles
are not seen on the brain surface, since they are located
inside, being turned into the cavity of the third ventricle
of the brain (Fig. 3c). In Cerebavis, the auditory tuber-
cles are as well developed as in extant birds; however,
they come onto the dorsal surface of the midbrain
(Fig. 3b). As a result, the auditory tubercles are located
dorsally, while the midbrain roof is located ventrally.
Thus, the basic difference of the brain pattern of Cere-
bavis from extant birds, Archaeopteryx, and extant rep-
tiles is the formation of the midbrain roof by the paired
auditory tubercles.
The well-developed quadrigeminal plate in the mid-
brain roof is a distinctive feature of mammals; however,
both pairs of their tubercles (auditory and optic) are
positioned along the longitudinal axis of the brain.
Cerebavis also shows both pairs of tubercles, but the
auditory tubercles are located dorsally, as in mammals,
while the optic tubercles are located ventrally, as in liv-
ing birds; this distinguishes Cerebavis from mammals.
A similar pattern of the midbrain roof was probably
characteristic of the dromaeosaurid Bambiraptor fein-
bergi Burnham et al., 2000 (see below), but has not
been recorded in any other extant or extinct vertebrates
examined in this respect.
The interbrain (diencephalon) of Cerebavis is only
represented by the epiphysis, which is clearly seen
between the auditory tubercles, caudal to the parietal
(parapineal) organ (Pl. 6, figs. 2, 4). The epiphysis was
rather large and was probably connected to the parietal
organ by an asymmetrical nerve. This pineal complex is
responsible for the circadian rhythms in the activity of
Cerebavis. Both structural elements are well devel-
oped, suggesting that Cerebavis was a twilight hunter.
It should be emphasized that, in the majority of verte-
brates, the epiphysis is located near the posterior edge
of the cerebral hemispheres, rostrally or caudally above
the epithalamus, or comes onto the surface of the mid-
brain. The epiphysis of Cerebavis is on the midbrain
roof. This position is characteristic of mammals and
atypical for reptiles (Saveliev, 2001).
The afterbrain (metencephalon) of Cerebavis is only
represented by small lateral fragments and the cavity of
the cerebellum, which is not preserved (Pl. 6, figs. 1, 5;
Pl. 7, figs. 13–15). The outline of this cavity show that
the cerebellum was extended dorsoventrally, as in the
majority of extant birds.
The Cenomanian beds of Melovatka could have
contained Hesperornis and Ichthyornis. Therefore, we
compared the brain of Cerebavis and these Late Creta-
ceous birds. Judging from the reconstructions produced
by Marsh (1880), the brain of Cerebavis clearly differs
from that Hesperornis regalis and Ichthyornis dispar in
the structure of the olfactory and optic lobes and mid-
brain roof. Marsh reconstructed the brain of Hesperor-
nis and Ichthyornis based on their skulls. He concluded
that the cerebral hemispheres of Hesperornis were
extended pear-shaped in outline, with elongated olfac-
tory lobes (Fig. 4a). The cerebellum was also elon-
gated, the optic lobes did not project out of the lateral
edges of the cerebral hemispheres. In Ichthyornis, the
cerebral hemispheres were more rounded in outline and
m
fcb
op
c
ol
mop col
(a)
(b)
Fig. 4. Reconstruction of the brain of Hesperornis and Ichthyornis (after Marsh, 1880): (a) Hesperornis regalis and (b) Ichthyornis
dispar. Designations: (c) cerebrum, (cb) cerebellum, (f) flocculi, (m) medulla oblongata, (ol) olfactory lobe, and (op) optic lobe.

662
PALEONTOLOGICAL JOURNAL Vol. 40 No. 6 2006
KUROCHKIN et al.
the olfactory lobes were shorter than in Hesperornis
(Fig. 4b). The cerebellum and optic lobes were similar
in shape to those of Hesperornis. Marsh concluded that,
in both toothed birds, the brain was in general more rep-
tilian than avian. Edinger (1951), on the contrary,
believed that the brain of Hesperornis and Ichthyornis
was more avian than reptilian, and criticized the recon-
structions proposed by Marsh, indicating that, in Hes-
perornis, only the interbrain (diencephalon) and the
region of the cerebellum and, in Ichthyornis, only the
olfactory lobes and a piece of the cerebrum could be
reconstructed. Jerison (1973), on the contrary, indi-
cated that Marsh produced outstanding reconstructions,
correctly demonstrating a true brain pattern of these
Cretaceous birds.
The Enaliornithidae were described from the Lower
Cretaceous (Albian) beds of Great Britain. The taxo-
nomic position of this group was the subject of consid-
erable discussion and different solutions. Galton and
Martin (2002) provided convincing arguments for their
affinity to Hesperornithiformes. In addition to numer-
ous bone fragments of the postcranial skeleton,
Enaliornis is represented by several caudal fragments
of the skull. Based on one fragment, Elzanowski and
Galton (1991) reconstructed the caudal part of the brain
of Enaliornis barretti. The caudal region of the cerebral
hemispheres, optic lobes, and myelencephalon of
Enaliornis are substantially dorsoventrally flattened.
The cerebral hemispheres are substantially smaller than
in extant Neornithes, while the optic lobes are displaced
under the myelencephalon; this essentially differs
between the brains of Cerebavis and Enaliornis.
Isolated endocranial molds of pterosaurs have also
been found (Edinger, 1941); in addition, a pterosaurian
brain was recently examined using computer tomogra-
phy (Witmer et al., 2002). Old and new data show that
the brain of pterosaurs is generally similar to that of
extant birds; however, the cerebral hemispheres are
only slightly larger in volume than the optic lobes, the
olfactory tracts and bulbs are short and small, while the
auditory tubercles of the quadrigeminal plate do not
come onto the midbrain roof. These features essentially
differ between the brains of Cerebavis and pterosaurs.
It is also expedient to compare Cerebavis with some
theropod dinosaurs. The most thorough information is
available on troodontids and the dromaeosaurid Bambi-
raptor. A brain cast of Troodon formosus Leidy, 1856
found in the Campanian of Canada is elongated antero-
posteriorly, and its cerebrum is even smaller than the
midbrain or afterbrain; it is noteworthy that the three
parts of the brain are equal in height (Currie and Zhao,
1993). The basiparasphenoid and parasphenoid of Tro-
odon formosus (as those of the other troodontid Sinove-
nator changi Xu et al., 2002) are narrow and project
ventrally (see Xu et al., 2002). Consequently, Cere-
bavis has nothing in common with troodontids in the
structure of the brain and the basal part of the skull.
Burnham (2004) reconstructed the brain of the dro-
maeosaurid Bambiraptor feinbergi from the Upper
Cretaceous of the United States, and also gave us his
unpublished manuscript on the brain of Bambiraptor.
The brain cast of Bambiraptor was produced based on
its braincase. The text is accompanied by stereophoto-
graphs of this cast and a drawing of the reconstructed
brain. The brain of C. cenomanica is substantially
smaller in absolute size than that of B. feinbergi. The
cerebral hemispheres of Cerebavis differ in the more
rounded outline and relatively larger volume. The
olfactory tracts of Bambiraptor are as thick as in Cere-
bavis, but shorter, while its olfactory bulbs were proba-
bly smaller. On the caudal side of the hemispheres of
Bambiraptor, after an abrupt ventral curvature, there is
the midbrain roof, with the auditory tubercles, which
occupy the same position as in Cerebavis. The optic
tubercles lie ventrolateral to the auditory tubercles, they
are relatively larger and more rounded than in Cere-
bavis. In our opinion, Burnham also mistakes the mid-
brain of Bambiraptor for the cerebellum. The brain cast
of Bambiraptor lacks a cerebellum. Thus, Cerebavis
and Bambiraptor are similar in the structure of the mid-
brain roof, while the cerebral hemispheres of Bambi-
raptor are less spherical and smaller in volume.
Cerebavis cenomanica Kurochkin et Saveliev, sp. nov.
Plate 6, figs. 1–6; Plate 7, figs. 1–17
Etymology. From the Cenomanian Stage.
Holotype. PIN, no. 5028/2, brain mold with
fragments of bony tissue and molds of bony cavities of
the auditory region of the skull and sphenoids; Russia,
Volgograd Region, Zhirnovskii District, 2 km east of
the village of Melovatka, Melovatka-3 locality; Upper
Cretaceous, Middle Cenomanian, Melovatskaya For-
mation, phosphorite horizon.
Description. The excellent preservation of the
surface structure of the fossil brain enables the recogni-
tion of the cerebral hemispheres, olfactory bulbs, pari-
etal organ, midbrain, epiphysis, and the cavity of the
cerebellum. The cerebral hemispheres are rounded oval
(pear-shaped), but voluminous and large. Rostrally,
they pass into thick olfactory tracts, which terminate in
extended bases of large olfactory bulbs. A distinct pari-
etal organ (pineal recess) is located on the caudal slope
of the hemispheres, in the depression of the interhemi-
spheric fissure. A voluminous midbrain is located cau-
dal to the hemispheres, its roof has a large epiphysis
between the auditory tubercles (Pl. 6, figs. 2, 4). The
optic tubercles in the shape of relatively small oval
prominences are located caudoventrally, under the
cerebral hemispheres. The cerebellum is only repre-
sented by its cavity, which is located just caudal to the
midbrain; judging from its structure, the cerebellum
was extended dorsoventrally and was not large (Pl. 6,
fig. 5; Pl. 7, figs. 16, 17). The x-ray sections show that
the surface of the brain mold is bordered by a dark thin
layer (tenths of millimeter thick), and its interior is
filled with almost homogeneous fine-grained matrix

PALEONTOLOGICAL JOURNAL Vol. 40 No. 6 2006
ON THE BRAIN OF A PRIMITIVE BIRD FROM THE UPPER CRETACEOUS 663
(Pl. 7, figs. 9–15). It is possible to recognize boundaries
between particular brain structures and even their inter-
nal elements. Of cranial bones, only the interorbital
septum and widely spaced spots of bony tissue on the
external surface and inside the mold are preserved
(Pl. 7, figs. 2–6, 14, 15). In addition, certain bone struc-
tures, such as the auditory region and the skull base, are
represented by molds filled with calcium phosphate and
silica (Pl. 6, 7).
The fossil brain enables the estimation of the gen-
eral outline of the skull of Cerebavis, with closely
spaced eyes, which are separated by a thin interorbital
septum and positioned ahead of the brain; this is char-
acteristic of a tropibasal skull unique to birds.
The basiparasphenoid is formed of a broad trapezoid
plate. The parasphenoid projects from under the basi-
parasphenoid and passes into the rostrum with a peculiar
swollen base. Two rostrolaterally directed projections are
located at the base of the parasphenoid, which are prob-
ably the bases of the basipterygoid processes.
Measurements, mm: greatest brain width at
the lateral boundaries of the cerebral hemispheres,
13.57; greatest depth of the specimen from the dorsal
surface of the cerebral hemispheres to the ventral sur-
Plate 6
TS
MS
BC
BS
TM
RS
IS
BO
TO
FI
RP EP TS
BC
T
TS
RT
T
IS
PS
BC
BC
BO
RS BC
BC
MT
C
BS
PS
RP RT
TM
FI
PO
EP
TS
C
MT
BS
TM
RT
T
FI
PO
RP
EP
TS MS BC
C
BC
1
23
4
5
6
MT
TM
Explanation of Plate 6
Figs. 1–6. Cerebavis cenomanica sp. nov., holotype PIN, no. 5028/2, brain mold; (1) left lateral, (2) dorsal, (3) right lateral,
(4) dorsolateral, (5) caudal, and (6) ventral views. Designations: (BC) bony structures filled with matrix, (BO) olfactory lobe,
(BS) basisphenoid, (C) cavity of the cerebellum, (EP) epiphysis cerebri, (FI) interhemispheric fissure, (IS) interorbital septum,
(MS) mesencephalon (midbrain), (MT) metencephalon (afterbrain), (PO) parietal organ, (PS) parasphenoid, (RP) pineal recess,
(RS) parasphenoid rostrum, (RT) telencephalon recess, (T) telencephalon (cerebrum), (TM) midbrain tectum (optic lobe),
(TO) olfactory tract, and (TS) torus semicircularis (auditory tubercles). Scale bar, 10 mm.

664
PALEONTOLOGICAL JOURNAL Vol. 40 No. 6 2006
KUROCHKIN et al.
face of the parasphenoid, 13.72; greatest length of the
specimen, 20.6; length of the olfactory tracts, including
the bases of the olfactory bulbs, ca. 8.5.
The volume of the fossil brain is estimated as ca.
1 cm3, i.e., it weighed ca. 1 g (Iwaniuk and Nelson,
2002). This volume includes some nonprepared pieces
of matrix preserved on the specimen, interorbital sep-
tum, and small portions of other bony elements, audi-
tory cavities, and skull base filled with phosphate–silica
substrate. Therefore, the brain could have been some-
Plate 7
1234
5678
2
7
11
15
TO
IS
RS
T
9
13
12
14 15
PS
TS
TM BS
16 17
10 11
C
Explanation of Plate 7
Figs. 1–17. Tomography of the brain and cranial fragments of Cerebavis cenomanica sp. nov., holotype PIN, no. 5028/2: (1) dorsal
view of the specimen, (2–15) cross sections, obtained with the aid of a Skyscan 1072 x-ray microtomograph, and (16, 17) stereopho-
tographs of the brain, caudolateral view (without olfactory bulbs). Black transverse lines and numbers in Fig. 1 designate sections
and their numbers in Figs. 2–15; for other designations, see Pl. 6. Scale bar: (1–15) 10 mm and (16, 17) out of scale.

PALEONTOLOGICAL JOURNAL Vol. 40 No. 6 2006
ON THE BRAIN OF A PRIMITIVE BIRD FROM THE UPPER CRETACEOUS 665
what smaller in volume and mass. However, the speci-
men lacks the afterbrain, medulla oblongata, and cere-
bellum. Thus, the estimate ca. 1 cm3 is probably close
to the actual volume.
R e m a r k s. The brain size of Cerebavis cenoman-
ica is similar to that of the London specimen of Archae-
opteryx lithographica. The greatest height of the brain
of Archaeopteryx is 10.0 mm, the width between the
lateral edges of the optic lobes is 14.5 mm (measured
using the figure from Alonso et al., 2004, text-fig. 3);
however, Jerison (1973) estimated the mean height of
the brain of the London specimen as 7.4 mm, and its
width as 7.2 mm. The volume of the brain of C. cen-
omanica is approximately 1 cm3. In the London speci-
men of Archaeopteryx, it was estimated as 1.5 cm3
(after Alonso et al., 2004) or 0.92 cm3 (after Jerison,
1973); the body mass was estimated as 500 g or 310–
425, respectively. It is difficult to estimate the body
mass of Cerebavis. When comparing in direct propor-
tion the size of its skull with living birds, the estimates
of body mass range from 30 g (based on passerines) to
80 g (based on nonpasserines) (Iwaniuk and Nelson,
2002; Alonso et al., 2004). However, this approach is
fraught with an error, since the brain and other systems
of Cerebavis were probably rather primitive. Neverthe-
less, taking into account the outline of the skull, which
is similar in many respects to living ornithurines, the
body mass of Cerebavis was probably at most twice
greater than 30–80 g. Consequently, the encephaliza-
tion index of Cerebavis is much greater than in Archae-
opteryx.
The volume of the endocranial cast of the brain of
Bambiraptor is 14 cm3, which corresponds to 12.6 g of
weight; the length of the brain is 55.2 mm, the height is
31.3 mm, and the greatest width through the midbrain
hemispheres (through the “cerebellum hemispheres”
after Burnham, 2004) is 27.5 mm; the body mass of
Bambiraptor is estimated as 1860–2240 g, with the
skeleton less than 50 cm high and about 100 cm long,
including the 35-cm-long tail (Burnham, 2004). Thus,
the relative size of the brain of Cerebavis was substan-
tially greater.
Anatomical features of the brain of C. cenomanica
turn us to the Enantiornithes. This group of advanced
feathered sauriurines developed functional adaptations
for active flapping flight through the skeletal structures
formed in parallel with ornithurine birds. These mor-
phological features of enantiornithines evolved from
the basis inherited from Archaeornithes; therefore, they
differed from ornithurines in many structural details of
the postcranial skeleton, although the general structural
pattern of the skeleton was very similar (Kurochkin,
2001, 2006). The Enantiornithines also had an elon-
gated rod-shaped coracoid, a saber-shaped scapula,
V-shaped furcula, bean-shaped humeral head, consoli-
dated carpometacarpus and metatarsus, complex verte-
brae, well-developed pygostyle, sternum with a keel,
etc. However, each element differed essentially in
structural details from ornithurine birds. Thus, the
Enantiornithes showed a high specialization to air,
ground, and aquatic locomotion, which was as pro-
found as in ornithurine birds, although it developed on
a different and relatively primitive background of their
skeletal system. The same is generally true of the brain
of C. cenomanica. We have shown that, among cur-
rently known extinct feathered diapsids, it could not
belong to taxa related to theropod dinosaurs, Archaeop-
teryx, Ichthyornis, Hesperornis, Enaliornis, or extant
Neornithes. Of all known feathered creatures, Cere-
bavis may be assigned, using indirect evidence, only to
Enantiornithes. However, relationships of Cerebavis
with primitive Early Cretaceous Ornithurae and some
feathered Mesozoic groups, such as Patagopterygidae,
Kuszholiidae, Zhyraornithidae, Confuciusornithidae,
and others remain an open question, because their brain
structure is uncertain.
Material. Holotype.
CONCLUSIONS
Cerebavis had a mosaic brain structure, combining
advanced characters of the central nervous system of
extant ornithurine birds with nonspecialized features of
reptiles. It was characterized by keen vision and exces-
sively specialized optic tubercles of the quadrigeminal
plate, which occupied the position typical for birds. It
also had a large telencephalon, which is comparable in
relative size to that of living birds, and a cerebellum
extended dorsoventrally. These features place it close to
Neornithes. However, the cerebrum of Cerebavis
remained pear-shaped and had thickened and elongated
olfactory tracts and large olfactory bulbs, which are by
no means characteristic of living birds. The sense of
smell was probably used for orientation, in searching
for food or, possibly, for mates during the breeding sea-
son. The large olfactory analyzer and pear-shaped cere-
bral hemispheres suggest an archaic brain construction
of Cerebavis.
It should be emphasized once again that the brain of
Cerebavis is characterized by unique features, which
occur neither in reptiles nor in extant birds. These are
primarily the well-pronounced auditory tubercles
(which are the torus semicircularis eminences) located
on the dorsal surface of the midbrain. The size and
shape of these tubercles suggest well developed hear-
ing. Cerebavis probably even had an external ear
formed of feathers. It may well have been able to
equally use eyesight, sense of smell, and hearing for
night hunting. The well-developed epiphysis and pari-
etal organ are evidence of adaptation to a nocturnal
mode of life. This complex controls daily hormonal
activity of the epiphysis depending on illumination.
The advanced control system of the circadian rhythms
in activity means that it played an important role. Tak-
ing into account the small body size, it is possible that
Cerebavis hid during the afternoon and hunted at night,
as is characteristic of many living rallids. This small bird

666
PALEONTOLOGICAL JOURNAL Vol. 40 No. 6 2006
KUROCHKIN et al.
could have been a generalist predator, feeding predomi-
nantly on various small land and aquatic invertebrates.
The unique brain design represented by Cerebavis
has not been repeated in the subsequent evolution of
vertebrates; however, it indicates the wide diversity of
evolutionary routes followed by feathered creatures.
ACKNOWLEDGMENTS
We are grateful to L. Martin and D. Burnham for
offering the opportunity to prepare the specimen in the
Museum of Natural History of the University of Kansas
(Lawrence, United States), to the workers of the Elec-
tron Microscopy Laboratory of PIN, where chemical
analysis of the specimen was performed, and to
G.T. Ushatinskaya and A.Yu. Rozanov (PIN) for con-
sulting on the rapid bacterial fossilization of soft
organic tissues. We thank D. Burnham for discussions
on the brain structure of extinct theropods and for pro-
viding us with his manuscript on the brain of Bambi-
raptor. We thank V.K. Golubev and E.K. Sytchevskaya
(PIN) for the identification of elasmobranchs and bony
fishes from the Melovatka localities, A.G. Olfer’ev for
consulting on the stratigraphy of the Upper Cretaceous
of the Volga Region, A. V. Mazin for participation in
producing photographs, and A. V. Lopatin for a perusal
of the manuscript and useful discussions.
This study was supported by the Russian Founda-
tion for Basic Research (project no. 04-04-48829) and
the Russian State Program for Support of Leading Sci-
entific Schools (project nos. NSh-1840.2003.4 and
NSh-6228.2006.4).
REFERENCES
1. S. S. Abyzov, M. Welsh, F. Weatall, et al., Bacterial
Paleontology (Paleontol. Inst. Ross. Akad. Nauk, Mos-
cow, 2002) [in Russian].
2. P. D. Alonso, A. C. Milner, R. A. Ketcham, et al. “The
Avian Nature of the Brain and Ear of Archaeopteryx,”
Nature 430 (7000), 666–669 (2004).
3. A. O. Averianov, E. N. Kurochkin, E. M. Pervushov, and
A. V. Ivanov, “Two Bone Fragments of Ornithocheiroid
Pterosaurs from the Cenomanian of Volgograd Region,
Southern Russia,” Acta Palaeontol. Polon. 50 (2), 251–
256 (2005).
4. G. R. de Beer, Archaeopteryx lithographica: A Study
Based upon the British Museum Specimen (Taylor Gar-
nett Evans and Co., Watford, 1954).
5. D. A. Burnham, “New Information on Bambiraptor fein-
bergi (Theropoda: Dromaeosauridae) from the Late Cre-
taceous of Montana,” in Feathered Dragons, Ed. by
P. J. Currie, E. B. Koppelhus, M. A. Shugar, and
J. L. Wright (Indiana Univ. Press, Bloomington, 2004),
pp. 67–111.
6. L. M. Chiappe and L. M. Witmer (eds.), Mesozoic Birds:
Above the Heads of Dinosaurs (Univ. California Press,
Berkeley, 2002).
7. P. J. Currie and X.-J. Zhao, “A New Troodontid (Dino-
sauria, Theropoda) Braincase from the Dinosaur Park
Formation (Campanian) of Alberta,” Can. J. Earth Sci.
30 (10–11), 2231–2247 (1993).
8. C. Dechaseaux, “Le cerveau d’Archaeopteryx est-il de
“type avien” ou de “type reptilien”?,” CR Acad. Sci.
Paris, Ser. D 267 (25), 2108–2110 (1968).
9. C. Dechaseaux, “Moulages endocraniens d’oiseaux de
l’Éocène supérieur du bassin de Paris,” Ann. Paléontol.
Vertébr. 56 (1), 69–72 (1970).
10. T. Edinger, “The Brain of Archaeopteryx,” Ann. Mag.
Natur. Hist., Ser. 9 18 (9), 151–156 (1926).
11. T. Edinger, “The Brain of Pterodactylus,” Am. J. Sci.
239 (9), 665–682 (1941).
12. T. Edinger, “The Brains of the Odontognathae,” Evolu-
tion 5 (1), 6–24 (1951).
13. A. Elzanowski and P. M. Galton, “Braincase of Enalior-
nis, an Early Cretaceous Bird from England,” J. Vertebr.
Paleontol. 11 (1), 90–107 (1991).
14. A. Elzanowski and P. Wellnhofer, “Cranial Morphology
of Archaeopteryx: Evidence from the Seventh Skeleton,”
J. Vertebr. Paleontol. 16 (1), 81–94 (1996).
15. P. M. Galton and L. D. Martin, “Postcranial Anatomy
and Systematics of Enaliornis Seeley, 1876, a Foot-pro-
pelled Diving Bird (Aves: Ornithurae: Hesperornithi-
formes) from the Early Cretaceous of England,” Rev.
Paléobiol. 21 (2), 489–538 (2002).
16. L. M. Gerasimenko, I. V. Goncharova, G. A. Zavarzin,
et al., “Dynamics of Release and Absorption of Phos-
phorus by Cyanobacteria,” in Ecosystem Rearrange-
ments and the Evolution of the Biosphere, Ed. by
A. Yu. Rozanov and M. A. Semikhatov (Nedra, Moscow,
1994), Vol. 1, pp. 348–353 [in Russian].
17. A. N. Iwaniuk and J. E. Nelson, “Can Endocranial Vol-
ume be Used as an Estimate of Brain Size in Birds?,”
Can. J. Zool. 80 (1), 16–23 (2002).
18. H. J. Jerison, “Brain Evolution and Archaeopteryx,”
Nature 219 (5161), 1381–1382 (1968).
19. H. J. Jerison, “Evolution of the Brain and Intelligence”
(Academic, New York, 1973).
20. E. N. Kurochkin, “Synopsis of Mesozoic Birds and
Early Evolution of Class Aves,” Archaeopteryx, No. 13,
47–66 (1995).
21. E. N. Kurochkin, A New Enantiornithid of the Mongo-
lian Late Cretaceous, and a General Appraisal of the
Infraclass Enantiornithes (Aves) (Palaeontol. Inst. Russ.
Acad. Sci., Moscow, 1996).
22. E. N. Kurochkin, “New Ideas about the Origin and Early
Evolution of Birds,” in Achievements and Problems of
Ornithology of Northern Eurasia at the Boundary of
Centuries, Ed. by E. N. Kurochkin and I.I. Rakhimov.
(Magarif, Kazan, 2001), pp. 68–96 [in Russian].
23. E. N. Kurochkin, “Parallel Evolution of Theropod Dino-
saurs and Birds,” Zool. Zh. 85 (3), 283–297 (2006).
24. E. N. Kurochkin, “New Fossil Birds from the Cretaceous
of Russia,” in Sixth International Meeting of the Society
of Avian Paleontology and Evolution: Abstracts, Quil-
lan, France, Ed. by E. Buffetaut and J. Le Loeuff (Quil-
lan, 2004), pp. 35–36.
25. O. Ch. Marsh, Odonthornithes: A Monograph on the
Extinct Toothed Birds of North America (Govern. Print.
Office, Washington, 1880).

PALEONTOLOGICAL JOURNAL Vol. 40 No. 6 2006
ON THE BRAIN OF A PRIMITIVE BIRD FROM THE UPPER CRETACEOUS 667
26. L. D. Martin, “The Origin and Early Radiation of Birds,”
in Perspectives in Ornithology, Ed. by A. H. Brush and
G. A. Clark (Cambridge Univ. Press, Cambridge, 1983),
pp. 291–338.
27. L. D. Martin, “A New Skeletal Model of Archaeopteryx,”
Archaeopteryx 13, 33–40 (1995).
28. J. Mlíkovsk , “Zwei Vogelgehirne aus dem Miozän
Böhmens,” asopis Mineral. Geol. 25 (4), 409–413
(1980).
29. R. Nieuwenhuys, The Central Nervous System of Verte-
brates (Spriger, Berlin–Heidelberg, 1998).
30. A. G. Olfer’ev and A. S. Alekseev, Stratigraphic Chart
of Upper Cretaceous Deposits of the East European
Platform: Explanatory Report (Paleontol. Inst. Ross.
Akad. Nauk, Moscow, 2005) [in Russian].
31. H. Osmólska, “Evidence on Relation of Brain to Endoc-
ranial Cavity in Oviraptorid Dinosaurs,” Acta Palaeon-
tol. Polon. 49 (2), 321–324 (2004).
32. E. M. Pervushov, M. S. Arkhangelsky, and A. V. Ivanov,
Catalogue of Localities with Remains of Marine Reptiles
in Jurassic and Cretaceous Deposits of the Lower Volga
Region (Kolledzh, Saratov, 1999a) [in Russian].
33. E. M. Pervushov, V. G. Ochev, A. V. Ivanov, and B. T. Yanin,
“Paleoecologostratinomic Characteristics of the Turo-
nian Phosphorite Horizon in the Vicinity of Zhirnovsk
(Volgograd Region),” in Selected Works of All-Russia Sci-
entific Conference on the Problems of Biosphere Studies
(Kolledzh, Saratov, 1999b), pp. 82–103 [in Russian].
34. S. V. Saveliev, Comparative Anatomy of the Vertebrate
Nervous System (Geotar-Med, Moscow, 2001) [in Rus-
sian].
35. S. V. Saveliev, The Origin of Brain (Vedi, Moscow,
2005) [in Russian].
36. K. N. Whetstone, “Braincase of Mesozoic Birds: 1. New
Preparation of the “London” Archaeopteryx,” J. Vertebr.
Paleontol. 2 (4), 439–452 (1983).
37. L. M. Witmer, S. Chatterjee, J. Franzosa, and T. Rowe,
“Neuroanatomy of Flying Reptiles and Implications for
Flight, Posture and Behaviour,” Nature 425 (6991), 950–
953 (2002).
38. X. Xu, M. A. Norell, X.-L. Wang, et al., “A Basal Tro-
odontid from the Early Cretaceous of China,” Nature 415
(6873), 780–784 (2002).
y
′
C
ˇ