VJ&r

P*x.

r*

A Lower Eocene Frigatebird

from the

Green River Formation of Wyoming
(Pelecaniformes: Fregatidae)

V

apt

'V*

KjLljl

V%

*

HEM

SVA

fe.

v%

Vis;-3

Ll|

STORRS L. OLSON

*>

sp*®*

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

NUMBER 35

aCpRPpfl

•**

SERIES PUBLICATIONS OF THE SMITHSONIAN INSTITUTION

Emphasis upon publication as a means of "diffusing knowledge” was expressed
by the first Secretary of the Smithsonian. In his formal plan for the Institution, Joseph
Henry outlined a program that included the following statement: "It is proposed to
publish a series of reports, giving an account of the new discoveries in science, and
of the changes made from year to year in all branches of knowledge.” This theme
of basic research has been adhered to through the years by thousands of titles issued
in series publications under the Smithsonian imprint, commencing with Smithsonian
Contributions to Knowledge in 1848 and continuing with the following active series:

Smithsonian Contributions to Anthropology
Smithsonian Contributions to Astrophysics
Smithsonian Contributions to Botany
Smithsonian Contributions to the Earth Sciences
Smithsonian Contributions to the Marine Sciences
Smithsonian Contributions to Paleobiology
Smithsonian Contributions to Zoology
Smithsonian Studies in Air and Space
Smithsonian Studies in History and Technology

In these series, the Institution publishes small papers and full-scale monographs
that report the research and collections of its various museums and bureaux or of
professional colleagues in the world cf science and scholarship. The publications are
distributed by mailing lists to libraries, universities, and similar institutions throughout
the world.

Papers or monographs submitted for series publication are received by the
Smithsonian Institution Press, subject to its own review for format and style, only
through departments of the various Smithsonian museums or bureaux, where the
manuscripts are given substantive review. Press requirements for manuscript and art
preparation are outlined on the inside back cover.

S. Dillon Ripley
Secretary

Smithsonian Institution

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY • NUMBER 35

A Lower Eocene Frigatebird

from the

Green River Formation of Wyoming
(Pelecaniformes: Fregatidae)

Starrs L. Olson

SMITHSONIAN INSTITUTION PRESS
City of Washington
1977

ABSTRACT

Olson, Storrs L. A Lower Eocene Frigatebird from the Green River Formation
of Wyoming (Pelecaniformes: Fregatidae). Smithsonian Contributions to Paleo¬
biology , number 35, 33 pages, 31 figures, 1977.—A new subfamily, Limnofrega-
tinae, is erected for the first known Tertiary specimens of the pelecaniform family
Fregatidae. These are described as a new genus and species, Limnofregata
azygosternon. The holotype is a nearly complete skeleton with feather impres¬
sions. Two additional specimens, consisting of a partial skeleton and a fragment
of ulna, are referred to the same species. All are from freshwater lake deposits
in the Lower Eocene Green River Formation of Wyoming, which are roughly
50 million years old. The principal differences between Limnofregata and modern
frigatebirds (Fregata) are seen in its shorter and less hooked rostrum, proportion¬
ately shorter wings and longer hindlimb, in the lack of fusion in the pectoral
girdle, and in the lack of extensive pneumatization of the skeleton. The fossil
form is, nevertheless, well advanced along the lines of modern frigatebirds and
nothing in its morphology seems to preclude its being ancestral to Fregata. The
fossil provides additional support for placing the suborder Fregatae between the
more primitive tropicbirds (suborder Phaethontes) and the members of the sub¬
order Pelecani. Frigatebirds in the early Tertiary evidently occupied different
habitats or a wider range of habitats than the modern forms and have only
subsequently been restricted to a purely oceanic environment.

Official publication date is handstamped in a limited number of initial copies and is recorded
in the Institution’s annual report, Smithsonian Year. Series cover design: The trilobite Phacops
ran a Green.

Library of Congress Cataloging in Publication Data
Olson, Storrs L.

A Lower Eocene frigatebird from the Green River formation of Wyoming (Pelecaniformes:
Fregatidae)

(Smithsonian contributions to paleobiology ; no. 35)

Bibliography: p.

1. Limnofregata azygosternon. 2. Paleontology—Eocene. 3. Paleontology—Wyoming. I. Ti¬
tle: A lower Eocene frigatebird from the Green River formation of Wyoming ... II. Series:
Smithsonian Institution. Smithsonian contributions to paleobiology ; no. 35.

QE701.S56 no. 35 [QE872.P4] 560'.8s [568'.4] 77-608097

Contents

Page

Introduction . 1

Acknowledgments. 2

Order Pelecaniformes Sharpe, 1891 . 2

Suborder Fregatae (Sharpe, 1891) . 2

Family Fregatidae Garrod, 1874 . 2

Subfamily Fregatinae, new rank. 2

Subfamily Limnofregatinae, new subfamily. 2

Limnofregata, new genus. 2

Limnofregala azygosternon, new species. 3

Description . 5

Skull and Mandible. 9

Presacral Vertebrae . 11

Pelvis . 15

Caudal Vertebrae. 15

Ribs . 17

Coracoid . 17

Scapula. 18

Furcula . 19

Sternum . 19

Humerus . 21

Ulna and Radius . 21

Free Carpals . 24

Carpometacarpus . 24

Phalanges of Manus . 24

Femur. 26

Tibiotarsus and Fibula. 26

Tarsometatarsus . 27

Toes . 29

Discussion . 29

Literature Cited. 33

Reconstruction of the
early Eocene frigatebird
Limiwjregata azygoslernon.

A Lower Eocene Frigatebird

from the

Green River Formation of Wyoming
(Pelecaniformes: Fregatidae)

Storrs L. Olson

Introduction

Among paleontologists it is often said that the
best collecting is to be done in museums. The truth
of this became quite evident to me in 1973, when in
the collections of the National Museum of Natural
History I encountered a nearly complete skeleton
of an unidentified bird from the Green River
Formation of Wyoming. This had been accessioned
in 1960 but had never been studied subsequently.
After a bit of chipping away at the matrix cover¬
ing critical portions of the skeleton, I was soon
convinced of the importance of this specimen, and
further preparation revealed it to be of consider¬
able significance in interpreting aspects of the evo¬
lution of the Pelecaniformes.

The modern members of the order Pelecani¬
formes are placed in three suborders and six
essentially monogeneric families: Phaethontes
(Phaethontidae, tropicbirds), Fregatae (Fregatidae,
frigatebirds), and Pelecani (Pelecanidae, pelicans:
Sulidae, gannets and boobies; Phalacrocoracidae,
cormorants; Anhingidae, anhingas). The first two
suborders are considered to be the most aberrant of
the order and it has at times been suggested that
each might be more closely related to some avian

Storrs L. Olson, Department of Vertebrate Zoology, National
Museum of Natural History, Smithsonian Institution, Wash¬
ington, D. C. 20560.

order other than the Pelecaniformes (see Sibley
and Ahlquist, 1972, for a summary of the history
of classification of Pelecaniformes).

Until now, there has been no fossil record of the
Fregatidae other than Pleistocene to Recent speci¬
mens of extant species of Fregaia found on St.
Helena (Olson, 1975) and Fernando de Noronha
(Olson, in prep.) in the South Atlantic. Thus, the
Eocene fossil that is the subject of this paper is of
great interest, for it is unquestionably a frigatebird.
Although it is a primitive frigatebird, within the
order it is sufficiently specialized to be regarded as
belonging to the same family as the modern genus
Fregaia. It is certainly as much a frigatebird as its
contemporary Hyracotherium (= “Eohippits”) is a
horse. Hyracotherium is placed in the family
Equidae, along with modern horses, but is main¬
tained in a separate subfamily. I propose a similar
treatment for the Eocene frigatebird.

It is often not generally appreciated that the
addition of a new genus to a monogeneric family
necessitates a redefinition of that family, and like¬
wise that the erection of a new subfamily within
such a family requires a diagnosis for the newly
formed nominate subfamily as well. Therefore, I
have briefly attempted such diagnoses in the follow¬
ing pages, these being based entirely on osteological
characters, inasmuch as these are all that are avail¬
able for the fossil form.

1

2

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Acknowledgments.— The careful preparation of
the holotype of the new species was undertaken by
Leroy Glenn, who spent many hours at this task.
I am grateful to the Division of Vertebrate Paleon¬
tology, National Museum of Natural History, Smith¬
sonian Institution, and particularly to Nicholas
Hotton III, for allowing this work to be done. I
thank Harrison Sheng and Daniel }. Stanley for
X-ray photographs of the specimen that facilitated
its preparation. I am indebted to Paul O. McGrew
for providing information and for photographs of
the University of Wyoming specimen, to Pierce
Brodkorb for calling it to my attention, and to Alan
Feduccia for lending a cast. Lloyd F. Gunther is
acknowledged for his kind donation of the para-
typic ulnar fragment. The drawings (except Figures
3, 6, and 8) are the work of Deborah R. Horner;
the life reconstruction was done by Anne Curtis;
and the photographs (except Figure 4) were taken
by Victor E. Krantz. I thank these people for help¬
ing to make the subject of this paper one of the
best illustrated of Tertiary birds. For comments and
criticisms that have much improved the manuscript
I am indebted to John Farrand, Jr., M. Ralph
Browning, Robert J. Emry, Frances C. James, Larry
D. Martin, and David W. Steadman.

Order PELECANIFORMES Sharpe, 1891

Suborder FREGATAE (Sharpe, 1891)

Included Family.— Fregatidae.

Diagnosis. —As for the family.

Family FREGATIDAE Garrod, 1874

Included Subfamilies. —Fregatinae, new rank;
Limnofregatinae, new subfamily.

Diagnosis.— Hindlimb more reduced than in
other Pelecaniformes. Tarsometatarsus proportion¬
ately the shortest of the order, less than 20% of
total hindlimb length (23% to 26% in other fam¬
ilies). Sternum short, width as great as or greater
than length (length greater than width in remainder
of order). Cervical vertebrae longer and narrower
than in Phaethon and lacking the hemal arches and
the abrupt change in structure within the series
that characterize the Pelecani. Deltoid crest of hu¬

merus large and triangular (unlike the Pelecani in
which this crest is greatly reduced) and not situated
as far distally as in the Phaethontes.

Subfamily FREGATINAE, new rank

Included Genus. —Fregata Lacepede, 1799.

Diagnosis.— Rostrum dorsoventrally flattened,
strongly hooked at tip. In adults, bony nostril open
only posteriorly. Rostrum twice as long as cranium.
Furcula fused both to sternum and to coracoids.
Scapular facet of coracoid convex; coracoidal articu¬
lation of scapula indistinct. Posterior border of
sternum between posterolateral processes entire. Pos¬
terior portion of pubis expanded. Humerus with
deltoid crest sharply triangular; bicipital surface
oval, inflated, and extending distally well down the
shaft. Wing long in proportion to body; hindlimb
short, tarsometatarsus greatly reduced, only about
15% of total length of hindlimb. Skeleton highly
pneumatized nearly throughout.

Subfamily LIMNOFREGATINAE, new subfamily

Included Genus. — Limnofregata, new genus.

Diagnosis.— Rostrum not markedly flattened dor¬
soventrally and only moderately hooked at tip.
Bony nostril open in a long slit. Rostrum only 1.5
times as long as cranium. Furcula entirely free from
sternum and coracoids. Scapular facet of coracoid
concave; coracoidal articulation of scapula promi¬
nent. Posterior border of sternum four-notched.
Pubis of nearly uniform width. Deltoid crest of
humerus not acutely pointed; bicipital surface and
crest reduced, not extending distally down the shaft
and not inflated. Wing about 12% shorter in pro¬
portion to body size than in Fregatinae; hindlimb
relatively longer, tarsometatarsus forming 19% of
total length of hindlimb. Skeleton not highly pneu¬
matic, apparently only the humerus and perhaps
the femur with small pneumatic openings.

Limnofregata, new genus

Type-Species.— Limnofregata azygosternon, new
species.

Diagnosis.—A s for the subfamily.

Etymology.— From the Greek limne (a marshy
lake) and the generic name Fregata (from Italian

NUMBER 35

3

fregata, a swift warship, a frigate; used as the ge¬
neric name of modern frigatebirds after that ap¬
plied in the vernacular by pre-Linnean authors who
were alluding to the superior flying ability and pi¬
ratical habits of these well-known seabirds). The
generic name of the fossil form is intended to refer
to the lacustrine habitat in which its remains were
interred.

Limnofregata azygosternon, new species

Holotype. —Nearly complete, partially flattened
skeleton with feather impressions, vertebrate pale¬
ontological collections of the National Museum

of Natural History, USNM 22753 (Figures 1, 2,
and others); collected in 1960 or previously by M. L.
Larsen.

Type-Locality. —“On the north side of Highway
30 North. At an elevation of 7250 ft [2250 m] near
the old diggings of Dave Haddenhams at Fossil,
[Lincoln County], Wyoming” (M. L. Larsen in litt.
to G. A. Cooper, 24 January 1961). This locality is
now within Fossil Butte National Monument.

Horizon. —Fossil Butte Member, Green River
Formation, late early Eocene.

Etymology. —From the Greek azygos (unyoked)
and sternon (breast or chest), in allusion to the lack

Figure 1.—Holotype (USNM 22753) of Limnofregata azygosternon , new genus and species, in
three pieces of slab and two of counterslab before any of the individual bones were removed
from the matrix.

4

of fusion in the elements of the pectoral girdle.
The specific name is a neuter noun in apposition.

Paratypes. —University of Wyoming 6919 (Fig¬
ure 4); partial skeleton; from SE 1/4, Sec 16, T21N,
R117W, Lincoln County, Wyoming; Fossil Butte
Member, Green River Formation, probably Lost
Cabinian, late early Eocene; collected by Carl
Ulrich.

USNM 243766 (Figure 22); proximal fourth of
right ulna; collected from the Green Riven Forma¬
tion near Kemmerer, Lincoln County, Wyoming,
by Lloyd F. Gunther.

Measurements. —Unless otherwise indicated, the
following measurements are taken from the holo-
type of Limnofregata azygosternon. All are in milli¬
meters.

Skull and Mandible. —Length of mandible 122.7;
length of rostrum from posterior margin of nostril
to tip 67.6; greatest depth of mandible (at surangu-
lar) 9.6; least depth of mandible 4.5; height of
quadrate from posterior portion of mandibular ar¬
ticulation to tip of otic process 15.3; width of inter¬
orbital bridge ca. 17; length of lacrimal attachment
12.2.

Vertebrae (measurements in parentheses are
from a small male of Fregata ariel, USNM 497972,
of the same body size as L. azygosternon ).—Height
of atlas from ventral margin of centrum to dorsal
margin of lateral process 9.0 (9.0); length of axis
from anterior surface of centrum to posterior mar¬
gin of hypapophysis 9.4 (8.1); length of 3rd cervical
from anterior margin of prezygapophysis to pos¬
terior margin of anapophysis 13.6 (13.0); length of
4th cervical taken in the same manner 15.0 (14.0);
length of centrum of 5th cervical 10.8 (12.0); length
of centrum of 8th cervical 11.8 (12.7); length of cen¬
trum of 10th cervical 11.4 (13.0); length of centrum
of 15th cervical 7.6 (10.0); length of centrum of 16th
vertebra 7.2 (9.9); length of centrum of 19th verte¬
bra 8.9 (9.2); height of 5th caudal 8.9 (9.9); width
across transverse processes of 6th caudal 14.8 (18);
height of 6th caudal 10.8 (11.7).

Pelvis.— Length of synsacrum 61.5; width across
anterior expansion of ilia 33.2; length of ilium ca.
82; width across antitrochanters ca. 40.

Coracoid.— Length from head to internal distal
angle 61.8; length of glenoid facet 12.0; length from
head to lowest point of glenoid facet 17.9; width of
sternal end ca. 25.

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Scapula (measurements in parentheses from cast
of University of Wyoming specimen).—Length 67.6
(67.7); width of proximal end 11.4 (11.6); width of
shaft at midpoint, [obscured in holotype] (5.0).

Furcula.— Length from anterior initiation of
coracoidal expansion to tip of hypocleidium 57.8.

Sternum. —Length from internal anastamosis of
sulci to xiphium ca. 45; width at level of 1st costal
facet 52 (estimated); distance from carinal apex to
xiphium 56.6; length from first to last costal facet

16.7.

Humerus.— Overall length ca. 140; proximal
width (above deltoid crest) 25.7; length of deltoid
crest 29.5; length of external condyle 10.6; approxi¬
mate distal width 18.5.

Ulna. —Length 166 (from cast of University of
Wyoming specimen). Depth of distal end ca. 12.5
(estimated from holotype). The following measure¬
ments are from the paratype USNM 243766: proxi¬
mal width 10.1; proximal depth 14.2; length of
prominence for anterior articular ligament 14.0.

Radius.— Depth of distal end ca. 9.5; shaft width
at midpoint 4.0.

Radiale.— Greatest diameter 10.5.

Ulnare. —Greatest diameter 11.7.

Carpometacarpus. —Length 80; length of inter-
metacarpal space 51.4; proximal depth ca. 17.5;
length of metacarpal I 10.2; depth of distal end

10 . 8 .

Phalanges of Manus. —Digit I, length 25.2,
proximal depth 5.5; digit II phalanx 1, length 39.8,
distal depth 10.1; digit II phalanx 2, length 33.3,
proximal depth 6.7; digit III, length 15.8, proximal
depth 4.4.

Femur. —Length ca. 55; distal depth ca. 10; trans¬
verse diameter of head 5.3.

Tibiotarsus.— Overall length 67; width through
condyles 8.4.

Fibula.— Proximal depth of head 3.6.

Tarsometatarsus. —Overall length 26.7; proxi¬
mal width 9.6; least width of shaft 5.6; width across
trochleae 11.3; width of middle trochlea 3.5.

Metatarsal I.—Length 10.4.

Phalanges of Pes (lengths). Digit I: phalanx 1,
15.1; phalanx 2, 8.1. Digit II: phalanx 1, 19.3; pha¬
lanx 2, 18.1; phalanx 3, 9.2. Digit III; phalanx 1,
19.3; phalanx 2, 19.8; phalanx 3, 18.0; phalanx 4,
lacking. Digit IV: phalanx 1, 12.2; phalanx 2, 11.0;
phalanx 3, 10.4; phalanx 4, 13.3; phalanx 5, 7.5.

NUMBER 35

5

Preservation and Preparation of the Speci¬
mens. —The greater part of the holotype skeleton
is preserved on a large slab of soft, cream-colored,
varved oil shale, now broken into three pieces, one
of the breaks being due to a natural cleavage in
the rock. Most of the skull and the sternum were
in smaller pieces of the counterslab (Figure 1). Nu¬
merous fragments of both slab and counterslab bear
feather impressions. All of the bones are a rich,
dark brown in color; the feather impressions are
black and are mostly indistinct except for the pri¬
maries of the right wing, the tips of which are
lacking.

As now viewed (Figures 1, 2), the specimen ap¬
pears to be lying more or less on its back, with the
right side uppermost. The pelvis faces ventral side
upward, and both wings extend out in front of
the body with their internal aspect upward. The
skull and mandible had been broken into four
pieces, but when these were united very little was
found to be missing, although the resulting speci¬
men is greatly flattened and difficult to interpret.
The neck extends forward over the right wing and
there appears to have been a postmortem separation
of the cervical vertebrae where they cross over the
right humerus. On the right wing, impressions of
the alula and some of the primaries are clearly pre¬
served. A section of the shafts of the radius and
ulna (restored with wax for the photographs) and
the proximal end of the carpometacarpus and the
carpals are missing from the right wing. Part of the
proximal end of the left humerus is covered by the
left coracoid, and most of its distal end (restored),
as well as the proximal ends of the left radius and
ulna, are missing; the distal elements of this wing,
however, are all present and well preserved. Bones
of the thoracic region are jumbled and overlap each
other considerably. The external surface of the right
coracoid is exposed, but the bone has been cleft by
a fault in the shale and is poorly preserved. Most
of the left coracoid is exposed in internal view,
being partly overlain by the furcula, ribs, and right
coracoid. Most of the furcula is exposed in anterior
view. Most of the right scapula is exposed, while
all but the posteriormost part of the left scapula is
obscured. The pelvis, with the last thoracic verte¬
bra, is clearly exposed in ventral view and overlaps
all but the head and the internal distal aspect of
the right femur and much of the trochanter of the

left femur. The tibiotarsi and tarsometatarsi on
both sides are seen in posterior view, whereas the
toes on both sides are seen in anterior view. It is
difficult to conceive of just how the tibiotarsi and
tarsometatarsi of this specimen could have turned
over completely on their longitudinal axes without
being displaced laterally or affecting the orienta¬
tion of the femora and toes. Three whole phalanges
and two halves are missing from the right foot and
the ungual phalanx of the left third toe is also lack¬
ing. Caudal vertebrae, perhaps as many as seven,
are scattered over and under the left foot and poste¬
rior portions of the pelvis. Unfortunately, the pygo-
style appears to be absent and there are no remain¬
ing impressions of tail feathers. Most of the sternum,
the distal portion of the left femur, and the proxi¬
mal portion of the left tibiotarsus came away with
a small piece of counterslab. The sternal carina re¬
mained on the main slab.

In preparing the holotype, all the portions of the
skull and mandible were united and entirely re¬
moved from the matrix. The right humerus and
right tibiotarsus, tarsometatarsus, and toes were like¬
wise freed in order to be able to study both sides
of the specimens. Also, the carina was removed from
the main slab and joined with the rest of the
sternum.

Of the paratype at the University of Wyoming
(Figure 4), I have studied an accurate cast and pho¬
tographs. The specimen consists of most of the right
wing, lacking only phalanx 2 and the distal half
of phalanx 1 of digit II. The humerus is in palmar
view, whereas the rest of the wing is seen from the
opposite side. Also present are most of the vertebral
column, ribs, and the right scapula, right coracoid
and the furcula in right lateral view. The specimen
is identical in size and all other respects to the
holotype and there is no question that the two
specimens are referable to the same species.

The paratypic proximal end of an ulna (Figure
22) would probably have been unidentifiable by
itself. When compared with the ulna of the holo¬
type, however, it was found to be morphologically
identical. The specimen is black in color and is
very little distorted by compression.

Description

In the illustrations and descriptions, Limnofre-
gata has been compared chiefly with Sula and

Figure 2. —Retouched photograph of the holotype of Limnofregata azygosternon (USNM 22753).
The skull, body of the sternum, distal end of the left femur, and proximal end of the right
tibiotarsus are shown in reverse in order to present the entire specimen associated as it was
originally preserved. (Scale = 40 mm)

NUMBER 35

7

Figure 3.—Diagram identifying various parts of the specimen shown in Figure 2. (Abbreviations
ending with l indicate the element is from the left side, while r indicates the right side; co =
coracoid, cp = carpometacarpus, cv = caudal vertebra, d1 = first digit, d2p1 = second digit
first phalanx, d2p 2 = second digit second phalanx, d3 = third digit, f = femur, fu = furcula,
H = humerus, mb = mandible, pe = pelvis, r = radius, rd = radiale, sc = scapula, sk = skull,
sp = sclerotic plates, st = sternum, tb = tibiotarsus, tm = tarsometatarsus, t1 = first toe,
t 2 = second toe, t 3 = third toe, t 4 = fourth toe, u = ulna, un = ulnare, vc = vertebral col¬
umn.)

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Figure 4— Partial skelelon of Limnojregata azygos lemon, paratype, University of Wyoming 6919,

(Abbrevations as in Figure 3.)

NUMBER 35

9

Fregata, the intention being to show how the two
genera of frigatebirds differ from each other and
how both differ from the Pelecani. Sula was selected
to represent the Pelecani, the Sulidae being the
most primitive family of that suborder and, there¬
fore, the most likely to share any characters with
frigatebirds. More extensive comparisons of Limno-
fregala with the Phaethontes are planned for a
later paper dealing with the fossil forms of that
group. For the time' being it can be said that
Limnofregata is very distinct from the Lower Eo¬
cene genus Prophaethon Andrews, for which Harri¬
son and Walker (1976) have recently created a new
order, Prophaethontiformes. Their interpretation
seems extreme and I believe that Prophaethon is
best retained in the Pelecaniformes, within which it
is on a line quite separate from Limnofregata.

Compared to modern frigatebirds, Limnofregata
is decidedly small. The pelvis and sternum are very
close in size to those of males (the smaller sex) of
the modern species Fregata ariel, the smallest living
member of the family. Therefore, F. ariel provides
a convenient standard for comparing the limb pro¬
portions of the fossil and modern forms (Figure
31), since all the species of Fregata have very simi¬
lar proportions. The following descriptions are
based on the holotype of L. azygosternon unless
otherwise stated. The degree of ossification of the
limb bones of the holotype indicates that it was an
adult, there being no sign of the porosity character¬
istic of juveniles. The three known individuals of
Limnofregata azygosternon are virtually the same
in size; either all three are of the same sex or there
was no sexual dimorphism in size in the fossil form.

Skull and Mandible

The skull and mandible have been considerably
flattened and are now essentially two-dimensional,
the dorsal and left lateral aspects being seen on
one side (Figures 5, and 6) and the ventral and
right lateral aspects on the other (Figures 7, 8). The
general shape of the bill is intermediate between
that of Sula and Fregata (Figure 9). The tip of the
rostrum is wider and not as deeply hooked as in
Fregata, and probably did not curve downward past
the tip of the mandible as in that genus. The tip
of the rostrum is only slightly hooked in Sula, being
most distinctly curved in Sula sula. The rostrum

of Limnofregata bears no evidence of having been
clorsoventrally compressed in life, as in Fregata,
and in this respect is more similar to Sula. Com¬
pared to Fregata, the cranium in Limnofregata is
proportionately larger and the bill shorter and
again is more similar to that of Sula (the length of
the rostrum is 1.5 times the length of the cranium
in Sula and Limnofregata, and 2 times as long in
Fregata).

Due to crushing it is difficult to determine the
full extent of the nostril in Limnofregata. The
posterior margin is distinct and lies a little less
than 10 mm in front of the nasofrontal hinge. The
nostril continues anteriorly as a well-defined open
slit for at least 30 mm, beyond which the two sides
have been pressed together, but it appears as though
the nostril actually continued out nearly to the
beginning of the hooked portion of the rostrum.
In adults of Fregata the nostril is open only pos¬
teriorly (<5 mm in F ariel, for example) and in
adults of the Sulidae it is closed altogether. In the
juvenile stages of these and other Pelecaniformes,
however, the nostrils are present as narrow open¬
ings running nearly the entire length of the rostrum
(Figure 9). These become occluded by bone in later
stages of development. Obviously, a long, open
nostril is the primitive condition and this appears
to have been retained in adults of the Eocene
frigatebird.

There are other indications that the skull of
Limnofregata was less ossified than in adults of
Fregata and more similar to juveniles of that genus.
In ventral aspect there is less ossification between
the anterior portions of the palatine and quad-
ratojugal bar. In lateral aspect there is less ossifica¬
tion in the space between the quadratojugal bar
and the nasal. The quadratojugal bar is straight and
narrow, slightly more expanded posteriorly. It is
quite similar to that of Fregata „ but proportion¬
ately heavier, and it lacks the great expansion in
the anterior portion seen in the Sulidae. It is
broader posteriorly than in Phaethon.

There are no grooves for salt glands on the dor¬
sal surface of the skull. If salt glands were present
in Limnofregata, they would have to have been in¬
side the orbit, as in all other Pelecaniformes. The
orbits are proportionately much larger than in Sula
or Fregata, their size being more like that of Pha¬
ethon, although the greater curvature of the dorsal

Figure 5 .—Dorsolateral view of the skull and mandible of the holotype of Limnojregata

azygosteriion. (Stereopair, natural size)

rim is more similar to Fregala. The orbital rims are
thin and turn upwards more than in Fregala. The
postorbital process is shorter and more blunt than
in Fregala or Phaelhon, but is unlike the short ex¬
panded process of Sula. The otic area, with the
quadrate articulation, is very similar to that of
Fregata and differs markedly from Phaelhon and
also from Sula, in which the opisthotic is directed
posteriorly rather than ventrally as in Fregata and
Limnofregala. The occipital condyle is larger and
more bulbous than in Fregata, in which the occipi¬
tal condyle is slightly notched and somewhat tri¬
angular. The occipital condyle is proportionately

much larger in Phaelhon and smaller in Sula than
in Limnofregala.

Except for part of the right palatine, most of
the palate is obscured. The palatine is long and
narrow, not inflated, and more closely resembles
the juvenile than the adult condition of Fregata.
Enough of the left pterygoid is visible to show that
it is similar to that of Fregata, though slightly
stouter, and unlike the long slender rod of Phaelhon
or the flat, expanded pterygoid of Sula. The lacri¬
mal is free and differs considerably from that of
Fregala and other known Pelecaniformes. The body
of the bone is long, oriented nearly horizontally,

NUMBER 35

11

Figure 6.—Diagram of the specimen shown in Figure 5 (at =
atlas, ax = axis, cv = cervical vertebrae, la = lacrimal,
mbl = left mandibular ramus, no = nostril, pop = postorbi¬
tal process, pt = pterygoid, qj = quadratojugal, qu = left
quadrate, sp = sclerotic plates) .

and is distinctly pointed both anteriorly and poste¬
riorly. In Fregata the body of the lacrimal is short,
oriented more vertically and is broad and rounded
posteriorly. In Limnofregata the descending process
of the lacrimal comes more from the posterior than
the anterior portion of the bone and appears to be
rather short and pointed. However, there is another
small piece of bone extending ventroposteriorly
from this process and abutting the quadratojugal
(Figure 5); this may represent part of the lacrimal

or possibly the os uncinatum. There is a thin elon¬
gate scar on the frontal for the attachment of the
lacrimal. The left quadrate is preserved upside-
down in the left orbit, with much of its internal
aspect visible on the left side of the specimen and
the articular socket for the quadratojugal protrud¬
ing on the other side near the right mandibular
articulation. It is distinctive in that the otic process
is very elongate. As in Fregata, the quadrate ap¬
pears almost single-headed, lacking the deep notch
between the heads that is typical of Sala.

The mandible is more similar to that of Fregata
than it is to any of the other members of the order.
It differs from Fregata as follows: dentary propor¬
tionately shorter and postdentary portion of mandi¬
ble longer: rami much deeper, not markedly nar¬
rowed as in Fregata ; mandibular symphysis shorter
and less deeply hooked; notch for articulation of
quadrate in lateral view noticeably deeper. The
different proportions of the mandible are in ac¬
cordance with the proportionately shorter rostrum
and longer cranium of Limnofregata.

Presacral Vertebrae

Although the vertebral column is imperfectly
preserved, a number of important features can be
discerned. In general, the presacral vertebral series
of Limnofregata is much more similar to that of
Fregata than to other Pelecaniformes. A positive
count cannot be made, but it is almost certain that
Limnofregata had the same number of vertebrae as
Fregata (15 cervicals and 4 thoracics). It did not
have the larger number of cervicals (17 to 20) found
in the higher pelecaniform families. The cervicals
are fairly elongate, as in Fregata, and are quite dis¬
tinct from the short, wide cervicals of Phaethon.

The atlas of Limnofregata is seen in left ventro¬
lateral view and is very similar to that of Fregata,
though slightly thicker anteroposteriorly, lacking
the expanded neural arch of Snla and being lower
and broader than in Phaethon. The axis can be
seen in left lateral and partial right lateral view
and is still articulated with the atlas. It differs most
notably from Sula in that the hypapophysis is not
an elongated, posteriorly directed spine. From
Phaethon it differs in its less developed neural
spine. It resembles the axis of Fregata but is propor¬
tionately larger and heavier, with the distance from

Figure 7.—Ventrolateral view of the skull and mandible of the holotype of
Limnofregata azygosternon. (Stereopair, natural size)

the anterior surface of the centrum to the begin¬
ning of the hypapophysis being greater, and the
neural spine being lower, directed more posteriorly,
and not being inflated.

The third cervical is partially disarticulated and
is displaced ventrally and laterally from the axis,
though it still presents its left lateral aspect. Like
the axis, this vertebra differs from Sula in the much
lesser development of the hypapophysis. It resem¬
bles that of Fregata and Phaethon but differs from
the former in having the styliform process much
smaller (although larger than in Phaethon ) and

from the latter in the larger, more inflated ana-
pophysis.

The fourth cervical is turned so that it is seen
principally in dorsal, and partly in ventral, view.
It is markedly different from that of Sula. In dorsal
view the notch between the postzygapophyses is
deeper and U-shaped, and it is quite different from
the very deep and V-shaped notch of Phaethon. In
ventral view, the articular surfaces of the postzyga¬
pophyses are seen to be markedly longer and nar¬
rower (4.0 X 1.8 mm) than in Fregata, in which
these surfaces are more rounded (3.4 X 2.3 mm in

NUMBER 35

13

Figure 8.—Diagram of the specimen shown in Figure 7
(al = articular surface for lacrimal, aq = quadratojugal
articulation of left quadrate, cv = cervical vertebrae, hy =
hyoid, mbl = left mandibular ramus, mbr = right mandib¬
ular ramus, oc = occipital condyle, op = opisthotic, pl =
palatine, qj = quadratojugal, sp = sclerotic plates, te =
right tomial edge of rostrum).

cf F. ariel). In Phaethon the postzygapophyses of
the fourth cervical are extremely attenuated, unlike
either genus of frigatebird.

The fifth cervical is completely detached from
the fourth and lies with its ventral side up in the
posterior corner of the orbit, the posterior end over¬
lapping the left mandibular articulation. A small
triangular piece of bone and matrix is missing from
this region in the holotype but it does not appear

extensive enough to have held another vertebra.
The fifth cervical of Limnofregata is longer and
narrower than the corresponding vertebra in Phae¬
thon or Sala and lacks the high neural crest of the
latter. It is very similar to that of Fregata but is
proportionately shorter, with longer styloid proc¬
esses and a deeper posterior surface of the centrum.

The remainder of the cervical series of the holo¬
type lies in a more or less articulated row, ventral
side uppermost. As mentioned, there appears to
have been a natural postmortem separation of the
neck where it crossed the right humerus. If so, the
next vertebra in the series is indeed the sixth. This
is further supported by counting anteriorly from
the reference point provided by the noticeable
transition between the 12th and the 13th cervicals,
which is present in both the modern and fossil
forms.

Vertebrae 6 through 9 are similar to each other
and, as far as can be seen, are very similar to those
of Fregata, the only difference noted being that the
articular surfaces of the postzygapophyses are pro¬
portionately larger in Limnofregata, while the over¬
all size of the vertebrae is smaller. In both genera
these vertebrae possess two well-developed cata-
pophyses. The anterior portion of the 9th cervical
is obscured in the holotype of Limnofregata, but
beginning with the 10th and continuing through
the 12th vertebra there is a fairly well-developed
semilunar hypapophyseal crest on the anterior half
of the centra. No such crests are present in Fregata;
the 10th vertebra still has two prominent cata-
pophyses, the 11 th has these much reduced and
placed closer together, and the 12th has only a
single small protuberance on the anteroventral sur¬
face of the centrum. Vertebrae 10—12 of Limnofre¬
gata are otherwise quite similar to those of Fregata.
Sula is very different in that vertebrae 9 through
13 have large hemal arches, a feature totally lacking
in either genus of frigatebird.

Only the centrum of the 13th cervical is clearly
discernable in Limnofregata and, as in Fregata, this
is noticeably shorter than that of the 12th cervical.
A low, thickened hypapophysis with a ridge run¬
ning the length of the centrum in Limnofregata is
nearly absent in Fregata.

Posterior to the 13th cervical is an area of crushed
bone that must represent the 14th cervical, but it
is too distorted to be interpreted. This is succeeded

Figure 9.—Skulls and mandibles in left lateral view: a, Sula sula; i$, Limnofregata azygoslernon
(the drawing is very generalized due to the difficulties in interpreting the crushed skull of the
holotype and is not to be relied on for details); c, juvenile Fregata minor showing the very long,
open nostril and the reduced ossification; d, adult Fregata arid. (Scale = 20 mm)

NUMBER 35

15

by a centrum in ventral view, no doubt that of the
15th cervical. The ventral surface of this centrum
is flat and ridged, as in the 15th and 16th vertebrae
of Fregata and unlike the centra of vertebrae 17
through 19, which are more rounded.

The next visible vertebra is likewise represented
by a centrum in ventral view. This is completely
disassociated from the anterior series and lies be¬
tween the right coracoid and scapula on the oppo¬
site side of the natural cleavage running through
the holotype slab. The ventral surface of this cen¬
trum is also flat, suggesting, by analogy to Fregata,
that this is the 16th vertebra.

Only two other presacral vertebrae are visible. If
the vertebral count is the same in Lirnnofregata as
in Fregata these would be by their positions cer¬
tainly the 19th and probably the 18th, suggesting
that it is the 17th vertebra that is missing or ob¬
scured. This is highly likely, since it is this area of
the specimen that has suffered the most destruction
by cleavage and that has the most overlap of bone.

What I am regarding as the 18th vertebra lies in
left lateral view amid a tangle of ribs and its orien¬
tation is reversed so that the anterior end points
more or less posteriorly. It is quite similar to the
corresponding vertebra of Fregata, with a well-
developed neural crest. The 19th vertebra is still
in articulation with the synsacrum and is seen in
ventral view. The barrel-like centrum is similar to
that of Fregata except that the margins of the artic¬
ulations form more distinct rims, giving the cen¬
trum slightly more of an hour-glass shape.

Pelvis

The pelvis (Figure 10) of Lirnnofregata is short,
wide, and shallow, unlike any of the Pelecaniformes
except Fregata and Phaethon. In the other families
of the order, the pelvis is long, laterally compressed
(particularly in the diving forms, such as cormo¬
rants) and deep (Figure 11). Lirnnofregata agrees
with Fregata and differs from Phaethon in the
greater expansion of the preacetabular ilia into
broad, rounded shields. In Fregata the pubis is
greatly attenuated in its anterior half, expanding
to nearly thrice its former width in the area of
contact with the ischium. In Lirnnofregata the
width of the pubis is about the same throughout
its length. In the synsacrum of Lirnnofregata the

transverse processes of the preantepenultimate ver¬
tebra are conspicuously larger than those of the
other vertebrae and extend laterally to abut against
the internal side of the antitrochanter, whereas in
Fregata the transverse processes are nearly uniform
from one vertebra to the next, and none extends
as far as the antitrochanter.

An interesting feature in Lirnnofregata is that
in the posterior margin of the innominate there is
a distinct notch and a resulting projection approxi¬
mately in the area of fusion of the ilium and
ischium. This is altogether absent in Fregata, but a
somewhat similar condition, though much less dis¬
tinct, is found in some individuals of Phaethon,
and a possibly homologous structure is prominent
in a more posterior position in Sula.

Caudal Vertebrae

In the holotype of Lirnnofregata I can account
for 6 and possibly 7 free caudal vertebrae, although
only four of these are not obscured by the pelvis
(Figure 10). In Fregata there may be either 6 or 7
free caudals, the first of these often fusing to the
synsacrum (Figure 11c). Of the exposed caudals in
Lirnnofregata, three can definitely be determined as
representing numbers 5, 6, and 7. The other is
probably the first or second free caudal. The caudals
of Lirnnofregata and Fregata are distinct from those
of Phaethon, in which the transverse processes are
much longer and narrower, or Sula, in which these
processes are expanded at their tips. In both Fregata
and Lirnnofregata the sixth and seventh caudals
differ from the remainder of the series in having
well-developed hemapophyses which are larger in
the seventh than in the sixth. The transverse proc¬
esses of the seventh caudal are better developed in
Lirnnofregata than in Fregata. The hemapophyses
of the sixth caudal in Lirnnofregata are slightly
shorter and axe well separated, rather than being
fused as in Fregata. All of the caudal vertebrae of
Lirnnofregata differ from those of Fregata in not
being inflated and, except for the 7th, in having
shorter and wider transverse processes. From this,
one can infer that the structure and shape of the
tail of Lirnnofregata was probably different from
that of Fregata. It was almost certainly not as long
or as deeply forked, and it may have appeared as
hypothesized in the frontispiece, though unfortu-

NUMBER 35

17

Figure 11. —Pelves in ventral view: a, Sula sula; b, Limnofregata azygosternon; c, Fregata ariel.

(Scale = 20 mm)

nately no impressions of the rectrices are preserved
in the holotype.

Ribs

The ribs of Limnofregata are generally similar to
those of Fregata but have decidedly longer uncinate
processes. These processes agree with Fregata and
differ from the Sulidae in being oriented more
nearly perpendicular to the shaft of the rib, rather
than being angled sharply dorsally.

Coracoid

Both coracoids of the holotype of Limnofregata
are damaged or partly obscured (Figure 12). Fortu¬
nately, the right coracoid, which is split by a fault
in the slab, is preserved with the ventral side up,
while the left is preserved with the dorsal side up.

In overall proportions, the coracoid of Limno¬
fregata is fairly similar to that of Phaethon. It is
longer and more slender than in Sula or Fregata
(Figure 13), but not so much as in Anhinga or
Phalacrocorax. The most striking feature is that

the coracoids of Limnofregata are not fused with
the furcula as is the case in Fregata. The area of
the furcular facet is considerably produced, how¬
ever, to an extent greater than in other Pelecani-
formes, although somewhat like Phaethon. The
glenoid facet is large with a well-developed rim,
unlike Phaethon, and is narrower and less rounded
than in Fregata, being more similar to Sula. In
ventral view the medial side of the sternal end
bears a wide, shallow depression that is lacking in
Sula, but present to a varying extent in Fregata and
also present in Phaethon, Anhinga, and Phalacro¬
corax.

The scapular facet in Limnofregata is distinctly
concave and cuplike, though not as deep or rounded
as in Pelecanus. In Phaethon there is only a slight
concavity and in the other modern genera of the
order, including Fregata, the scapular facet is essen¬
tially convex and indistinct. The procoracoid proc¬
ess is short, rounded, and imperforate, unlike the
thin, broad, perforate, bladelike procoracoid of
Phaethon or the elongate, pointed process of Sula
and Fregata.

The ventral sternal facet in Limnofregata is a

18

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Figure 12- —Furcula, left coracoid (dorsal aspect), sternal end of right coracoid (ventral aspect),
and proximal end of left humerus of the holotype of Limnofregata azygosternon. (Stereopair,
natural size)

low ridge on the medial fourth of the sternal end
of the coracoid, much better developed than in
Phaethon but not as deep as in Fregata and wider

Figure 13. —Left coracoids in dorsal aspect: a, Sula sula; b,
Limnofregata azygosternon; c, Fregata ariel. (Scale = 20 mm)

than in Sula. The dorsal sternal facet in Limno¬
fregata is quite distinct from that of Fregata, pre¬
senting a deep face situated high on the shaft, as
in Sula, but not situated as close to the medial mar¬
gin of the shaft. This facet in Phaethon is similar
in position to that of Limnofregata, but is not as
deep.

Scapula

The scapula of Limnofregata (Figure 14) is pro¬
portionately longer than in Fregata. The acromion
is more weakly developed than in any modern pele-
caniform but is more similar to that of Phaethon
ox Fregata than to the greatly elongated and pointed
acromion in other members of the order (Figure 15).
The coracoidal articulation is more prominent in
Limnofregata than in any modern pelecaniform
genus except Pelecanus, in which the coracoidal
articulation is a large, distinct ball. This corre¬
sponds with the deeper scapular facet of the cora¬
coid in both of these genera. The blade of the

NUMBER 35

19

Figure 14.—Right scapula (dorsal aspect) and the scapular
end of the right coracoid (ventral aspect) of the holotype of
Limnofregata azygosternon. (Stereopair, natural size)

scapula is thin and flat and appears to be of a more
uniform width than that of Fregata, Sula, or Pha-
ethon, the tip being less acutely pointed than in
those genera.

Furcula

The ventral face of the furcula is exposed in the
holotype of Limnofregata (Figure 12), with all of
the anterior articulation obliterated on the right
side and the scapular tuberosity obscured on the
left. In the University of Wyoming specimen, most
of the right ramus is visible in lateral view. Com¬
parison of the furcula of Limnofregata with that of
Fregata is rendered difficult because in the latter
this bone is broadly fused posteriorly with the ster¬
num and anteriorly with the coracoids, whereas
there is no such fusion in Limnofregata. The fur¬
cula in Limnofregata is in the shape of a broad V,
with the arms not as divergent and U-shaped as is
the case in the Sulidae, Anhingidae, Phalacrocora-
cidae, and Pelecanidae. There is no indication of

the strong rounded coracoidal facets seen in those
families, although there is a distinct expansion
below the scapular tuberosity. The scapular tuber¬
osity appears to have been rather elongate and
pointed as is also evident in Fregata, in spite of
the fusion. The rami are broad and deep and there
is a small, pointed hypocleidium but no evidence
of a large sternal articulation or of the small
pointed epicleidium seen in Fregata and Sula. The
furcula of Phaethon is simple and U-shaped with
almost no modification of the anterior ends. Despite
the fact that it is not fused to the rest of the pec¬
toral girdle, the furcula of Limnofregata appears
to be more similar to that of Fregata than to other
members of the order.

Sternum

The sternum in the holotype of Limnofregata is
well preserved and is seen in the ventrolateral as¬
pect of the left side (Figure 16). The body of the

Figure 15.—Right scapulae in dorsal view: a, Sula sula; b,
Limnofregata azygosternon; c, Fregata ariel. (Scale = 20 mm)

20

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Figure 16. —Ventrolateral view of the sternum of the holotype of Limnofregata azygosternon.

(Stereopair, about natural size)

sternum is very short, so that the width is equal
to or exceeds the length, a condition found in mod¬
ern Pelecaniformes only in Fregata (Figure 17). The
carina extends posteriorly the full length of the
bone, which is also characteristic only of Fregala,
although Phaethon is close. The carina in lateral
view is not as rounded as in Fregata. The carinal
apex is distinctly bifurcate and the only modern
pelecaniform showing a tendency towards such a

condition is Phaethon, in which, however, the bi¬
furcation is indistinct and lies anterior to a large
round furcular facet, a feature not found in Limno¬
fregata. The carinal apex of the Lower Eocene pele¬
caniform Prophaethon shrubsolei is also distinctly
bifurcated (Harrison and Walker, 1976:12, fig. 6c),
but this, too, has a large furcular facet as in Phae¬
thon. In Fregata this area is entirely obliterated
through fusion with the furcula.

Figure 17.—Sterna in ventrolateral view: a, Sula sula; b, Limnofregata azygosternon; c, Fregata

ariel. (Scale = 20 mm)

NUMBER 35

21

The sterno-coracoidal processes are slender and
pointed in Limnofregata, but not as much as in
Fregata, and differ from Sula and Phaethon, in
which the sterno-coracoidal processes are short and
triangular. Phaethon, Fregata, and Limnofregata
have six costal facets on the sternum, whereas there
are five in all the Pelecani except Anhinga, which
has only four. The posterior margin of the sternum
in Limnofregata is definitely 4-notched, the lateral
notch on each side being rather deep and the pos¬
terior lateral processes being slender and elongate.
Phaethon also has a 4-notched sternum but in this
genus the medial notches are the larger and the
lateral ones are very reduced, sometimes to different
degrees on either side of the same individual. In
all other pelecaniform genera the sternum is broadly
2-notched and in Fregata the xiphial area does not
form a projection, so that the posterior border be¬
tween the posterior lateral processes is entire.

The sternum in Limnofregata is decidedly closer
to that of Fregata than to any other modern genus
of the order. The 4-notched posterior border and
the lack of fusion with the furcula are primitive
characters, while the very short, broad shape ap¬
pears to be a derived condition.

Humerus

The humerus of Limnofregata (Figure 18) resem¬
bles that of Phaethon and Fregata and differs from
that of any of the Pelecani in having a large tri¬
angular deltoid crest. The shaft appears to have
been relatively stouter than in Fregata or Phaethon
but this may in part be due to distortion of the
fossil specimens.

The humerus of Limnofregata is in many respects
more similar to that of Phaethon than Fregata, the
principal differences from Phaethon being that the
deltoid crest is of a very different shape (Figure 19),
the bicipital crest is much less prominent, and the
tricipital fossa appears to have been smaller. Al¬
though much crushed, in the right humerus of the
holotype of L. azygoslernon it can be seen that the
tricipital fossa was pneumatic, and what appear to
be small trabeculae can still be discerned in what
is left of the foramen.

The humerus in Fregata presents a number of
striking specializations as compared to Limnofre¬
gata (Figures 19, 20). To begin with, it is very

highly pneumatic, with foramina evident in several
places in the tricipital fossa, bicipital furrow, brach¬
ial depression, and olecranal fossa. The bicipital
crest is inflated and bulbous and extends distally
for a considerable length, whereas in Limnofregata
the bicipital crest is all but lacking. The deltoid
crest in Fregata has a distinctly projecting apex
whereas in Limnofregata the apex is broadly
rounded. The attachment for M. pectoralis major
on the palmar surface of the deltoid crest of Limno¬
fregata has several protuberances and rugosities that
are more complex than in Fregata, although the
most distal and best developed of these appears to
be homologous to one in a similar position in Fre¬
gata. No other modern pelecaniform has such a dis¬
tinct attachment for M. pectoralis major. Both
Limnofregata and Fregata have a well-defined oval
scar for M. latissimus dorsi posterioris on the mid¬
line of the anconal surface of the shaft at the level
of the apex of the deltoid crest. This scar is less
distinct in Phaethon and is situated off the midline,
more towards the deltoid crest, as is the condition
in the other modern members of the Pelecaniformes
in which this scar is at all apparent.

The distal end of the humerus in Limnofregata
differs considerably from that of Fregata and is
more similar to that of Phaethon or even Sula.
Fregata differs in possessing a greatly inflated, elon¬
gate ridge extending proximally from the entepi-
condylar prominence, in the much greater develop¬
ment of the ectepicondylar area, and in the more
bulbous and inflated external condyle, which lacks
the slight hook at the proximal end seen in Limno¬
fregata. Limnofregata differs from Phaethon and
Sula and is closer to Fregata in its much greater
attachment of the anterior articular ligament and
the greater development of the ectepicondylar area.
The brachial depression is larger and deeper than
in Phaethon but not nearly as extensive as in Sula.

Ulna and Radius

As does the humerus, the ulna of Limnofregata
in most details bears more resemblance to that of
Phaethon than Fregata. The papillae for the secon¬
daries are not nearly as well developed as in Fregata
(Figure 21), although it appears that the internal
surface of the shaft was probably more rounded,
as in Fregata, rather than being somewhat flattened

c

Figure 18.—.Stereoviews of the right humerus of the holotype of Limnofregata azygoslernon:
A, anconal aspect (X 7/10); n, palmar aspect (x 7/10); c, distal end in palmar aspect (about X 1.3).

as in Phaethon. The most distinctive feature of the
ulna of Limnofregata is the very large, roughly tri¬
angular prominence for the anterior articular liga¬
ment (Figure 22), which has a tapering extension
along the lower margin of the impression of M.
brachialis anticus. This is most similar to the con¬

dition seen in Phaethon and very different from the
reduced and more rounded prominence of Fregata
or that of any of the other modern Pelecaniformes.
The impression of M. brachialis anticus is deeper
than in either Fregata or Phaethon, but like the
latter has no pneumatic foramina, in contrast to

NUMBER 35

23

K'fe'ijft

Figure 19.—Proximal ends of humeri in anconal view: a,
Phaethon aetherem; b, Limnofregata azygosternon; c, Fregata
ariel. (Scale = 20 mm)

mm

ri

Figure 20.— Humeri in palmar view: a, Sula sula; b, Limno¬
fregata azygosternon; c, Fregata ariel, (Scale = 20 mm)

Fregata. The olecranon is narrower and projects
farther proximally than in Fregata or Phaethon
and is more sharply set off from the internal cotyla.
The proximal radial depression is not as deep as
in Fregata but the tendinal attachment along its

Figure 21.—Ulnae in internal view: a, Sula sula; b, Limno¬
fregata azygosternon; c, Fregata ariel. (Scale = 20 mm)

24

r f

Figure 22-—Proximal end of right ulna of Limnofregata
azygosternon, paratype, USNM 243766. (Stereopair, about
X 1-8)

anterior margin is oriented nearly parallel to the
long axis of the shaft, as in Fregata, rather than
nearly perpendicular, as in Phaethon. The distal
end of the ulna in Limnofregata is fairly similar to
that of Fregata and differs markedly from Phaethon,
in which the carpal tuberosity is large and hooked
and bears a deep groove between it and the internal
condyle.

The radius of Limnofregata is more sharply tri¬
angular in cross section than in Fregata and is thus
more like Phaethon. The distal end is not greatly
inflated as in Fregata, although the ligamental
prominence is decidedly larger than in Phaethon,
as it is also in Fregata.

Free Carpals

The left radiale of the holotype of L. azygoster¬
non is seen in its proximo-internal aspect. It is more
similar to that of Sula and Fregata than to Phae¬
thon, in which last the medial surface between the
articulating facets is narrower. In Limnofregata this
surface lacks the pneumatic foramina that are pres¬
ent in Fregata.

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

The left ulnare of the holotype is seen mainly in
its internal aspect and this bears a tendinal groove
that is much deeper than in either Fregata or Sula
but which is similar to that seen in Phaethon.

Carpometacarpus

The carpometacarpus of Limnofregata (Figure 23)
resembles that of Fregata in having metacarpals II
and III parallel throughout their lengths rather
than diverging clistally as in Phaethon and Sula.
Metacarpal I is lower than in Fregata and the pisi¬
form process is less bulbous and does not project
as far outward; in these respects Limnofregata is
closer to Phaethon. The distal metacarpal symphy¬
sis is longer in Limnofregata than in Sula or Pha¬
ethon and the tuberosity of metacarpal II is not
crescentic nor as high as in Phaethon. In these re¬
spects, Limnofregata is fairly similar to Fregata
(Figure 24). There are no pneumatic foramina (at
least not on the internal face) in the carpometa¬
carpus of Limnofregata, unlike Fregata, in which
such foramina occur in several places.

Phalanges of Manus

Digit I of Limnofregata is not greatly expanded
proximally as it is in Fregata and is fairly similar
to that of Phaethon, but it is not as attenuated dis-
tally and is proportionately shorter (33% of length
of carpometacarpus in both genera of frigatebirds,
vs. 40% in Phaethon).

In phalanx 1 of digit II of Limnofregata the
distal margin is truncate, whereas in Phaethon there
is a clistally projecting ventral lobe and in Fregata
there is a large dorsal projection as well. In both
the right and left elements of the holotype of L.
azygosternon there is an irregular fenestra in the
distal third of the blade, as in Fregata.

Phalanx 2 of digit II is more similar to that of
Fregata than to the other genera compared. Both
possess a small foramen on the clorso-internal corner
of the proximal end. This is pneumatic in Fregata
but appears to have been a nutrient foramen in
Limnofregata. The proximal end is not as expanded
and projects more distinctly from the shaft in
Limnofregata than in Fregata. The distal end bears
a hooked ventral process, as in Fregata and Phae¬
thon, but the remaining distal portion is much

NUMBER 35

25

Figure 23.—Distal portions of right and left wings of the holotype of Limnofregata azygosternon.

shorter than in either genus and is not greatly at¬
tenuated as in Phaethon.

Digit III differs markedly from that of Phaethon,
which is very elongate and slender, extending over

half the length of digit II. This element is rather
similar to that of Fregata but it is proportionately
much shorter and more closely resembles that of
Snla.

26

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Figure 24.—Carpometacarpi in internal view: Sula sula; b, Limnofregata azygosternon; c,

Fregata ariel. (Scale = 20 mm)

Femur

The femora of the holotype of Limnofregata azy¬
gosternon are somewhat difficult to interpret since
both are partially obscured. Nevertheless, a number
of important features can be discerned. The femur
of Limnofregata differs markedly from that of Phae-
thon as follows: shaft much stouter; head flattened
proximally, not spherical; trochanter higher than
head; fibular groove much deeper. In having the
head flattened and lying below the level of the
trochanter, Limnofregata differs from Phaethon and
all of the Pelecani and agrees only with Fregata
(Figure 25). The femur of Limnofregata differs from
that of Fregata in being proportionately longer, in
having the attachment for the ligamentum teres
deeper and more distinct, and in the somewhat
greater development of the fibular condyle. In the
left femur of the holotype there is barely visible
the outline of what may have been a small pneu¬
matic foramen in the trochanter, now mostly ob¬
scured by crushing and by the overlapping of the

pelvis. The femur is pneumatic in both Fregata and
Sula but not in other modern Pelecaniformes.

Tibiotarsus and Fibula

The tibiotarsus of Limnofregata (Figure 26) is
somewhat longer proportionately than in Fregata
and has a decidedly heavier shaft than either Fre¬
gata or Phaethon. The fibula is likewise heavier
than in either of these genera and extends farther
distally and apparently was fused at its distal ex¬
tremity, as in Sula. In its overall build the tibio¬
tarsus of Limnofregata is more similar to Sula than
to other Pelecaniformes (Figure 27), but it differs
in a number of details, many of which are more
similar to Fregata.

The inner cnemial crest in Limnofregata is larger
and projects farther anteriorly than in Phaethon or
Fregata, in both of which this crest is greatly re¬
duced. Although folded over in the best preserved
of the two tibiotarsi of L. azygosternon, the inner
cnemial crest appears to have extended farther ante-

NUMBER 35

27

A B

Figure 25.—Femora in medial view: A, Sula sula; b, Limno-
fregata azygosternon; c, Fregata ariel. (Scale = 20 mm)

riorly even than in Sula and was thinner and more
bladelike. Its proximal extent, however, is not as
great as in Sula and in this respect it more closely
resembles Fregata. The outer cnemial crest, too, is
well developed and unciform, being less reduced
than in any of the modern Pelecaniformes except
Phalacrocorax and Anhinga, which are otherwise
quite different. The head of the fibula is relatively
and absolutely smaller than in Fregata or Sula.

Figure 26. —Right tibiotarsus of the holotype of Limnofre-
gata azygosternon , anterior view. (Stereopair, natural size)

Figure 27. —Tibiotarsi in anterior view: a, Sula sula; b, Lim-
nofregata azygosternon; c, Fregata ariel. (Scale = 20 mm)

The distal end of the tibiotarsus of Limnofregata
is similar to that of Fregata and differs from the
other genera compared in having a distinct internal
ligamental process. In anterior view both the inter¬
nal and external condyles in Limnofregata are
roughly triangular in shape, the apex of the tri¬
angles pointing towards the midline. This shape is
still evident in Fregata, although somewhat modi¬
fied, but in the other genera compared the condyles
in anterior view are oblong and of nearly uniform
width. The distal opening of the tendinal groove
is rounded as in Fregata and Phaethon, and is not
a transverse oval as in Sula. In posterior view the
crest of the internal condyle is better developed
than in Sula or Phaethon and more closely resem¬
bles the condition in Fregata. The anterior inter¬
condylar fossa is narrower in Limnofregata than in
Sula, but is wider than in Fregata and is not too
different from that of Phaethon.

T ARSOMETATARSUS

The tarsometatarsus of Limnofregata (Figure 28)
is quite distinctive. It bears little resemblance to

28

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

A

B

Figure 28. —Right tarsometatarsus of the holotype of Limnofregata azygosternon: a, anterior view;

b, posterior view. (Stereopairs, X 2)

the peculiar tarsometatarsus of Phciethon, which is
slender and has a very deep anterior groove. In its
overall proportions the tarsometatarsus of Limno¬
fregata is intermediate between the greatly reduced
and “degenerate” tarsometatarsus of Fregala and
the more elongate element of Sula (Figure 29). It
difEers from Sula and agrees with Fregata in having
the external cotyla projecting laterally beyond the
line of the external margin of the shaft and in hav¬
ing a lower and less pointed intercotylar promi¬
nence. The cotylae are deeper in Limnofregata than
in Fregata and the internal cotyla is proportionately
smaller and does not extend as far anteriorly. The
anterior metatarsal groove is fairly deep, and the
various tendinal tubercles and foramina are devel¬
oped about as in Sula, whereas in Fregata the ante¬
rior face of the bone is much modified, the most
conspicuous feature being the ridges and grooves,
which indicate the positions of the individual meta¬
tarsal elements, and the medial displacement of the
tubercle for the tibialis anticus.

The hypotarsus of Limnofregata is quite different
from that of either Sula or Fregata and actually
bears more resemblance to that of Phaethon. In
Sula the hypotarsus has a very large, posteriorly di¬
rected inner calcaneal ridge and contains two com¬
pletely enclosed canals. In Fregala the hypotarsus
is an irregularly shaped block containing a single
large enclosed canal. In Limnofregata there are no
closed canals and the hypotarsus consists of three
calcaneal ridges separated by two grooves, the inner
calcaneal ridge being the largest and swollen, the

ABC

Figure 29. Tarsometatarsi in anterior (top) and posterior
( bottom ) views: a, Sula sula; b, Limnofregata azygosternon;
c, Fregata ariel. (Scale = 20 mm)

NUMBER 35

29

two outer ridges being thinner and equally devel¬
oped. A similar condition exists in Phaethon except
that the outer ridge is the largest and most swollen
and the middle ridge is noticeably smaller than the
other two.

The distal foramen is well developed and circular
in Limnofregata and exits on the posterior surface
of the shaft, in all respects being similar to Sula
though relatively smaller. In Fregata the distal fora¬
men seen in anterior view is greatly reduced and
it exits between trochleae III and IV so that it is
not visible in posterior view.

The trochleae of Limnofregata in anterior view
may be characterized as follows: external trochlea
longer and more slender than in Fregata, more like
Sula but with the posterior wing less distinct; mid¬
dle trochlea much longer, more deeply grooved, and
not projecting sharply anteriorly from the shaft as
in Fregata, more similar to Sula but longer and
angling laterally instead of medially; internal troch¬
lea about as in Fregata, not as long and slender as
in Sula and not projecting as far medially. In poste¬
rior view the most notable feature of the trochleae
of Limnofregata is that the articulating surface of
the middle trochlea extends proximally in an elon¬
gate triangular shape as in Sula and is not short
and truncate as in Fregata. Both genera of frigate-
birds and Sula differ from Phaethon in having the
internal rather than the middle trochlea extending
farthest distally.

Toes

The “toe formula” in Limnofregata, going from
longest to shortest, is 3—4—2—1. Fregata and Pele-
canus have the same formula, as does Phaethon,
except that in the last-named genus digits 4 and 2
are about equal. In all other Pelecaniformes the
fourth toe is longer and either equals (Sula) or
greatly exceeds (Phalacrocorax and Anhinga) the
middle toe in length, the formula being 4 = 3-2-1,
or 4-3-2-1.

The toes of Limnofregata (Figure 30), while pro¬
portionately longer, are nevertheless much more
similar to those of Fregata than to those of Sula or
Phaethon in that they are considerably heavier and
have the phalanges of digit IV much shorter. The
phalanges of Limnofregata differ from Fregata in
that they are longer, more slender, and the articu¬

lating ends are much less expanded relative to the
width of the shafts; the ungual phalanges are less
curved.

Discussion

Limnofregata affords some new information on
the affinities of the Fregatidae, since it presents us
with a frigatebird without many of the obfuscating
specializations of Fregata. The following are some
of the characters that are shared by the Phaethon-
tidae and Fregatidae but not found in the families
of the suborder Pelecani: 15 cervical vertebrae, all
lacking hemal arches and changing only gradually
in shape through the series (see Mivart, 1878); ab¬
sence of coracoidal facets on furcula (not ascertain¬
able for the Fregatidae until now); 6 costal facets
on sternum (versus 5 or 4 in the Pelecani); humerus
with a large, triangular deltoid crest; width of pel¬
vis across anti trochanters greater than 80% of
length of dorsal portion of ilium (25-65% in Pele¬
cani). With the possible exception of the short,
broad pelvis, which may have evolved independ¬
ently in each of these two aerially adapted families,
all of the above characters appears to be primitive
and thu9 are not evidence for a close relation¬
ship between tropicbirds and frigatebirds. They do,
however, definitely support the phylogeny advo¬
cated by Lanham (1947), who treated the Phaethon-
tes and Fregatae as being primitive within the
order, to be followed by the more specialized fami¬
lies of the Pelecani. This is in contrast to such clas¬
sifications as those of Peters (1931) and Wetmore
(1960) where the frigatebirds are placed in a termi¬
nal position, separated from the Phaethontes by the
Pelecani—a sequence that cannot be justified by
the facts available now.

Limnofregata and Phaethon share some characters
that are not seen in Fregata: the proportions of the
coracoid, certain aspects of the humerus, the con¬
formation of much of the proximal end of the ulna,
the nature of the hypotarsus, and the 4-notched
sternum. The last-named is without doubt a primi¬
tive character and most or all of the others could
be also. Collectively, however, they might be taken
as a possible indication that frigatebird morphology
could have been derived from that of some primi¬
tive phaethontid-like ancestor.

In only a few characters can Limnofregata be

30

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Figure 30. —Tarsometatarsi and toes in anterior view: a, Sula sula; b, Limnofregala azygosternon;

c, Fregala ariel. (Scale = 20 mm)

said to resemble the Pelecani (particularly Sula)
more than either Phaethon or Fregata. The ratio
of cranium to bill length appears to be like that of
Sula but this hardly appears to be of significance,
since the condition in Limnofregala is intermediate
between that of Phaelhon and Prophaelhon and
that of Fregala. The tibiotarsus and tarsometatarsus
of Limnofregala resemble Sula rather closely in sev¬
eral respects. This, however, could be attributable
to the fact that these bones in both Phaelhon and
Fregala are highly modified, whereas the Sulidae,
which appear to stem from near the base of the
Pelecani, could be expected to retain a rather primi¬
tive morphology in the hindlimb. Thus, while the
conformation of the tarsometatarsus and tibiotarsus
cannot be cited as showing a close relationship be¬
tween Limnofregala and the Sulidae, it does sup¬
port their allocation to the same order.

Despite its primitiveness, the skeleton of Limno-
fregata has many more similarities to Fregala than

to any other pelecaniform genus. The conformation
of the otic area, pterygoid, mandible, cervical verte¬
brae, carpometacarpus, femur, and toes; the great
reduction of the hindlimb; the proportions of the
sternum and the general structure of the furcula;
the size and position of the scar for M. latissimus
dorsi posterioris and some other details of the hu¬
merus all resemble Fregala more closely than Phae¬
thon or any of the Pelecani.

Although at first Limnofregala seems very differ¬
ent from modern frigatebirds, on closer examina¬
tion the differences are actually rather less signif¬
icant than might be expected (Figure 31) in view
of the relatively great age of the fossil (ca. 50 mil¬
lion years bp). The proportionately longer, more
dorsally compressed and strongly hooked bill of
modern frigatebirds probably reflects increased spe¬
cialization for feeding from the surface of the water
and perhaps for taking marine animals, such as
squid, rather than freshwater fish. Possibly corre-

NUMBER 35

31

Figure 31.—Outline comparing the overall skeletal proportions of modern and fossil frigatebirds.
Fregata is depicted on the left half of this composite individual and Limnojregata is pictured
on the right. The numbers refer to the actual length measurements (mm) of individual bones
in the holotype of Limnojregata azygosternon and in a small male of Fregata ariel (USNM
497972) . Note that the body size of the two species is virtually the same, as evidenced by the
lengths of the sterna and pelves. (Scale = 100 mm)

Iated with this is the increased ossification of the
skull in Fregata. This may represent a general evo¬
lutionary trend in various seabirds; three skulls of a
Miocene albatross from Oregon, as in Limnofregata,
are more similar to juveniles of modern species than
to adults (Olson, notes).

Although the hindlimb of Fregata is more re¬
duced than in Limnofregata, that of the latter shows
that by the early Eocene the hindlimb of frigate-
birds had already become more shortened than in
other Pelecaniformes. Other specializations of Fre¬
gata appear to be correlated with an increasingly
aerial existence. These include lengthening of the
wing, fusion in the pectoral girdle, and extensive
pneumatization of the skeleton. This pneumatiza-
tion has in some instances resulted in fairly dra¬
matic changes in morphology compared to Limno¬
fregata, the most extreme example being the
humerus. The resultant morphological changes, how¬
ever, represent specializations peculiar to Fregata
and are departures from all other Pelecaniformes.
The fact that they also differ from Limnofregata is
not of particular concern in assessing the relation¬
ships of the fossil. There is nothing that I can de¬
tect in the skeleton of Limnofregata that precludes
its being directly ancestral to Fregata. The fact that
by the early Eocene it was already markedly special¬
ized along much the same lines as the modern genus
renders this possibility plausible.

The completeness of the material of Limnofregata
prompts me to make some observations and specula¬
tions about the taxonomic conclusions that might
have been drawn had only parts of its skeleton been
discovered and studied. The pelvis and probably
the sternum are the only elements of Limnofregata
that by themselves would most likely be recognized
as being similar to those elements in the Fregatidae.
The skull and possibly the tarsometatarsus, taken
in their entirety but alone, might also have been
properly assigned by an alert paleontologist. The
humerus, radius, ulna, carpometacarpus, coracoid,
furcula, scapula, tibiotarsus and probably the fe¬
mur, if found in isolation, either whole or in part,
would have stood very little change of being cor¬
rectly identified as once having formed a part of a
frigatebird, and almost any one of them might have
been said to constitute a new family, although it is
far from certain that in every instance each speci¬
men would have been assigned to the proper order.
The distal end of the tarsometatarsus or that of the
tibiotarsus might well have been referred to the
Sulidae, whereas the proximal end of the tarsometa¬
tarsus would have proved impossible to assign to a
modern family. The inadvisability of basing higher
taxonomic categories of Paleogene birds on frag¬
mentary limb elements, particularly when these bear
no close resemblance to modern taxa, thus becomes
evident.

32

SMITHSONIAN CONTRIBUTIONS TO PALEOBIOLOGY

Modern frigatebirds are confined to tropical
oceans and breed exclusively on islands. The known
environment of Limnofregata, on the other hand,
was one of expansive inland freshwater lakes con¬
taining a variety of fish that were subject to pe¬
riodic catastrophic die-offs (McGrew and Casilliano,
[1976]). From its structure one may conjecture that
Limnofregata probably occupied a niche somewhat
similar to that of modern gulls of the genus Laras.
It was better suited for flapping flight than is Fre-
gat a and was probably able to alight on and take
off from the surface of the water or even the shore,
unlike Fregata, for which the proportions of its
limbs present nearly insurmountable obstacles to
such activity. Limnofregata was no doubt a preda¬
tor and scavenger on the multitudes of clupeiform
fish found in the Green River lakes. It is possible

that it preyed upon chicks of the flamingo-like
wader Presbyornis pervetus, which nested in huge
colonies around the early Eocene lakes of Wyoming,
Colorado, and Utah (McGrew and Feduccia, 1973;
Feduccia and McGrew, 1974).

It appears that frigatebirds were not entirely, if
at all, marine in the early Tertiary, and this may
perhaps have been true of other groups that are
now entirely pelagic. Quite possibly the frigatebirds,
sulids, tropicbirds, Procellariiformes, and certain
primitive terns, such as Anous, may originally have
been more diverse ecologically but were subse¬
quently replaced in continental habitats by more
advanced groups, with the result that they have
been restricted to a purely oceanic environment
where they now exist in a sense as relicts.

Literature Cited

Feduccia, Alan, and Paul O. McGrew

1974. A Flamingolike Wader from the Eocene of Wyoming.
Contributions to Geology (University of Wyoming),
13(2):49-61, 13 figures.

Harrison, Colin James Oliver, and Cyril Alexander Walker
1976. A Reappraisal of Prophaethon shrubsolei Andrews
(Aves). Bulletin of the British Museum (Natural
History), Geology, 27(1): 1-30, 3 plates, 7 figures, 2
tables.

Lanham, Urless N.

1947. Notes on the Phylogeny of the Pelecaniformes. Auk,
64(l):65-70, 1 figure.

McGrew, Paul O., and Michael Casilliano
[1976.] The Geological History of Fossil Butte National
Monument and Fossil Basin. National Park Service
Occasional Paper, 3:1-37, 29 figures, 2 tables.

McGrew, Paul O., and Alan Feduccia

1973. A Preliminary Report on a Nesting Colony of
Eocene Birds. Pages 163-164 in E. M. Schnell, editor,
Wyoming Geological Association 25th Conference
Guidebook. 4 figures. Casper, Wyoming: Prairie
Publishing Company.

Mivart, St. George

1878. On the Axial Skeleton of the Pelecanidae. Trans-
actions of the Zoological Society of London, 10(7:1):
315-378, plates 55-61.

Olson, Storrs L.

1975. Paleornithology of St. Helena Island, South Atlantic
Ocean. Smithsonian Contributions to Paleobiology,
23:1^19, 6 plates, 10 figures, 8 tables.

In prep. Terrestrial Vertebrates of Fernando de Noronha
Island, South Atlantic Ocean.

Peters, James Lee

1931. Check-list of Birds of the World. Volume 1, 345
pages. Cambridge, Massachusetts: Harvard Uni¬
versity' Press.

Sibley, Charles G., and Jon E. Ahlquist

1972. A Comparative Study of the Egg White Proteins
of Non-Passerine Birds. Peabody Museum of Natural
History, Yale University, Bulletin, 39:1-276, 37 fig¬
ures.

Wetmore, Alexander

1960. A Classification for the Birds of the World. Smith¬
sonian Miscellaneous Collections, 139(11): 1—37.

33

REQUIREMENTS FOR SMITHSONIAN SERIES PUBLICATION

Manuscripts intended for series publication receive substantive review within their
originating Smithsonian museums or offices and are submitted to the Smithsonian Institution
Press with approval of the appropriate museum authority on Form SI-36. Requests for
special treatment—use of color, foldouts, casebound covers, etc. — require, on the same
form, the added approval of designated committees or museum directors.

Review of manuscripts and art by the Press for requirements of series format and style,
completeness and clarity of copy, and arrangement of all material, as outlined below, will
govern, within the judgment of the Press, acceptance or rejection of the manuscripts and art.

Copy must be typewritten, double-spaced, on one side of standard white bond paper,
with 114 " margins, submitted as ribbon copy (not carbon or xerox), in loose sheets (not
stapled or bound), and accompanied by original art. Minimum acceptable length is 30 pages.

Front matter (preceding the text) should include: title page with only title and author
and no other information, abstract page with author/title/series/etc., following the establish¬
ed format, table of contents with indents reflecting the heads and structure of the paper.

First page of text should carry the title and author at the top of the page and an unnum¬
bered footnote at the bottom consisting of author's name and professional mailing address.

Center heads of whatever level should be typed with initial caps of major words, with
extra space above and below the head, but with no other preparation (such as all caps or
underline). Run-in paragraph heads should use period/dashes or colons as necessary.

Tabulations within text (lists of data, often in parallel columns) can be typed on the text
page where they occur, but they should not contain rules or formal, numbered table heads.

Formal tables (numbered, with table heads, boxheads, stubs, rules) should be sub¬
mitted as camera copy, but the author must contact the series section of the Press for edito¬
rial attention and preparation assistance before final typing of this matter.

Taxonomic keys in natural history papers should use the alined-couplet form in the
zoology and paleobiology series and the multi-level indent form in the botany series. If
cross-referencing is required between key and text, do not include page references within the
key, but number the keyed-out taxa with their corresponding heads in the text.

Synonymy in the zoology and paleobiology series must use the short form (taxon,
author, year:page), with a full reference at the end of the paper under ‘‘Literature Cited."
For the botany series, the long form (taxon, author, abbreviated journal or book title, volume,
page, year, with no reference in the ‘‘Literature Cited") is optional.

Footnotes, when few in number, whether annotative or bibliographic, should be typed
at the bottom of the text page on which the reference occurs. Extensive notes must appear at
the end of the text in a notes section. If bibliographic footnotes are required, use the short
form (author/brief title/page) with the full reference in the bibliography.

Text-reference system (author/year/page within the text, with the full reference in a
"Literature Cited" at the end of the text) must be used in place of bibliographic footnotes in
all scientific series and is strongly recommended in the history and technology series:
“(Jones, 1910:122)” or . . Jones (1910:122).”

Bibliography, depending upon use, is termed "References,” "Selected References,” or
"Literature Cited.” Spell out book, journal, and article titles, using initial caps in all major
words. For capitalization of titles in foreign languages, follow the national practice of each
language. Underline (for italics) book and journal titles. Use the colon-parentheses system
for volume/number/page citations: "10(2):5-9.” For alinement and arrangement of
elements, follow the format of the series for which the manuscript is intended.

Legends for illustrations must not be attached to the art nor included within the text but
must be submitted at the end of the manuscript — with as many legends typed, double¬
spaced, to a page as convenient.

Illustrations must not be included within the manuscript but must be submitted sepa¬
rately as original art (not copies). All illustrations (photographs, line drawings, maps, etc.)
can be intermixed throughout the printed text. They should be termed Figures and should
be numbered consecutively. If several "figures” are treated as components of a single larger
figure, they should be designated by lowercase italic letters (underlined in copy) on the illus¬
tration, in the legend, and in text references: “Figure 9b,” If illustrations are intended to be
printed separately on coated stock following the text, they should be termed Plates and any
components should be lettered as in figures: “Plate 9b,” Keys to any symbols within an
illustration should appear on the art and not in the legend.

A few points of style: (1) Do not use periods after such abbreviations as “mm, ft,
yds, USNM, NNE, AM, BC.” (2) Use hyphens in spelled-out fractions: “two-thirds.” (3)
Spell out numbers “one” through “nine” in expository text, but use numerals in all other
cases if possible. (4) Use the metric system of measurement, where possible, instead of
the English system. (5) Use the decimal system, where possible, in place of fractions.

(6) Use day/month/year sequence for dates: “9 April 1976.” (7) For months in tabular list¬
ings or data sections, use three-letter abbreviations with no periods: “Jan, Mar, Jun,” etc.

Arrange and paginate sequentially EVERY sheet of manuscript — including ALL front
matter and ALL legends, etc., at the back of the text—in the following order: (1) title page,
(2) abstract, (3) table of contents, (4) foreword and/or preface, (5) text, (6) appendixes,

(7) notes, (8) glossary, (9) bibliography, (10) index, (11) legends.

mmfm-

mmm'

'■'sy/'y

Wwi*

/ $$$& -

: .v»» •••.•• vsaBovv vy/M